• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

iPhone 8 Is World's Fastest Phone (It's Not Even Close)

Fhtagn

Member
Right, I was arguing this earlier. I just don't see whats so remarkable here.

Given that the A series has been about a year ahead or more in single threaded performance for three or four years now, if the A11 isn't remarkable, the rest of the industry is garbage. Either Apple is outperforming everyone, or Apple is unremarkable and the rest are badly underperforming.
 

MThanded

I Was There! Official L Receiver 2/12/2016
This thread is sort of embarrassing. It's embarrassing for Apple fanboys who are clearly bothered that people don't give a shit about synthetic benchmarks because some Android owners they feel their phone is fast enough and reliably does everything they need. It's also embarrassing for Android fans that try compare old iPhones to newer Android phones in speed tests/benchmarks thinking there's barely any actual performance difference. It's especially sad for those who throw in the RAM specs for their support that Android phones are objectively better.


I'm glad that Apple keeps pushing the envelope though to at least keep the other OEMs on their feet. Google may be the first to respond if they decide to start designing their own in house chip.



Here we go again with this.

Maybe this article will give some insight. I'm honestly shocked that no here seems to have shared this or similar articles, but fanboys gotta fan and won't apply critical thinking.

http://www.pcworld.com/article/3006...he-ipad-pro-really-isnt-as-fast-a-laptop.html


And yes I'm aware that Apple's CPUs have made significant gains since then and they offer incredible performance. That still doesn't mean that one benchmark should be relied for everything...

This drive-by fanboy this and fanboy that post is embarrassing.
 

rambis

Banned
Given that the A series has been about a year ahead or more in single threaded performance for three or four years now, if the A11 isn't remarkable, the rest of the industry is garbage. Either Apple is outperforming everyone, or Apple is unremarkable and the rest are badly underperforming.
Lol this is hilarious. Ok dude.
 

Two Words

Member
Lol this is hilarious. Ok dude.
I am twice as fast as you. I am either considered a fast human being or an average human being. If I am fast, you are average. If I am average, you are slow. How is this a strange concept to you. We can’t both be average or both be fast if I am twice as fast as you.
 

rambis

Banned
How is being consistently more than a year ahead and growing not remarkable? The rest of the market isn't even close.

Because its not? I don't understand how you expect anybody to take you seriously when you only try to limit everything to one specific benchmark and ignore everything else.
 

Fhtagn

Member
Because its not? I don't understand how you expect anybody to take you seriously when you only try to limit everything to one specific benchmark and ignore everything else.

Ok, can you point me to a debunking of the general consensus in tech reporting that the A11 is way ahead of everyone else? Or a debunking of Geekbench's validity as a metric?

What is it that I am ignoring that suddenly makes the A11 average?
 

BaasRed

Banned
Apple surely would be tempted to bring the chips Macs to inhouse control as well, rather than let Intel's schedule dictate when they update.

Yes! Its to their advantage in every way. They have to only put some work in and the benefits are tremendous.
 

rambis

Banned
Ok, can you point me to a debunking of the general consensus in tech reporting that the A11 is way ahead of everyone else? Or a debunking of Geekbench's validity as a metric?

What is it that I am ignoring that suddenly makes the A11 average?
Geekbench only measures a particular set of synethic scenarios, mostly cpu dependent. There are far better real world benchmarks out there.

http://www.antutu.com/en/ranking/rank1.htm


What you are ignoring is that this is a rather tepid jump from the previous generations in every area except multicore performance. You seem to be caught in the new SOC smell but the leapfrogging of the competitors has been happening a while. This 3-4 yeat advantage you keep talking about is make believe.
 

mrklaw

MrArseFace
Geekbench only measures a particular set of synethic scenarios, mostly cpu dependent. There are far better real world benchmarks out there.

http://www.antutu.com/en/ranking/rank1.htm


What you are ignoring is that this is a rather tepid jump from the previous generations in every area except multicore performance. You seem to be caught in the new SOC smell but the leapfrogging of the competitors has been happening a while.

And on that one the Iphone 8 seems to hit over 200,000. Still average?

http://www.antutu.com/en/doc/111124.htm
 

Two Words

Member
Geekbench only measures a particular set of synethic scenarios, mostly cpu dependent. There are far better real world benchmarks out there.

http://www.antutu.com/en/ranking/rank1.htm


What you are ignoring is that this is a rather tepid jump from the previous generations in every area except multicore performance. You seem to be caught in the new SOC smell but the leapfrogging of the competitors has been happening a while.
A CPU benchmark SHOULD be CPU dependent. And that site doesn’t have the iPhone 8. The S8 and the iPhone 7+ are basically even even though the S8 is new and the 7+ is a year old in this test.
 

Theonik

Member
Phone are becoming exceedingly more powerful each year. Though I mostly just use my phone for texting and some apps like reddit and etc.

Not sure what people do with their phone that demand high performance.
That concept is nonsense tbh. Faster phones are faster at all given tasks so you get a really nice overall QOL improvement.
 

Future

Member
I’m on the upgrade plan, and I was debating not bothering upgrading because the 7 plus is so good. This is almost making me change my mind. Everyone wants a phone that’s gonna last a long time if possible, and this power will help it a long
 

Fhtagn

Member
What you are ignoring is that this is a rather tepid jump from the previous generations in every area except multicore performance. You seem to be caught in the new SOC smell but the leapfrogging of the competitors has been happening a while. This 3-4 yeat advantage you keep talking about is make believe.

I never said they were 3-4 years ahead, let alone keep saying it.

I'm not caught up in anything, I've been trying to figure out what point you're making in the context of how the rest of the market is struggling to keep up. Even if Apple's delta between generations is average from here out, that keeps them a year ahead.

And going from a chip that is two big / two small cores in an either/or configuration to a chip that is two big, four small, use all six at once as the OS sees fit, is actually a big jump. Especially for what this implies for the A11X, which will be used in iPads that support using three apps plus one picture in picture video at the same time.
 

borghe

Loves the Greater Toronto Area
And on that one the Iphone 8 seems to hit over 200,000. Still average?

http://www.antutu.com/en/doc/111124.htm
Hey look it’s an android user pointing to synthetic benchmarks that scale linearly with core counts and offscreen or compute results.

This goes back to what everyone was saying earlier on this page. Let’s look at real world processing use cases, like video editing where Note8 was over 3 mins, S8 was over 4 mins, and 8 was at 42 seconds. Because when it comes to synthetic benchmarks, Apple has more processing efficient CPUs while android typically has more cores. Thus iOS advocates will point to overall benchmarks where android advocates will point to screen unrestrained core scaling linear benchmarks. At the end of the day though all that matters is game FPS and computationally expensive workflow speeds.
 

The Real Abed

Perma-Junior
As much as I love my 7+, every day I find it harder and harder to not go crazy and buy an 8+ to replace it even though I got it in April.

But I can do it. I can do it. I don't need an 8+. I want it but I don't need it. I can hold off a year or two. Yeah. I can do that.
 
Geekbench only measures a particular set of synethic scenarios, mostly cpu dependent. There are far better real world benchmarks out there.

http://www.antutu.com/en/ranking/rank1.htm

So now Geekbench is CPU dependant, but before you were arguing that resolution had a large effect on Geekbench scores. Which is it?

Why is Antutu better? It seems to be another synthetic benchmark, so why do you seem to refer to it as a real world benchmark? It doesn't even split out single core and multicore performance. Why is it better to obfuscate that info?

Why does a $350 iPhone SE with a two year old SOC pants a $750 Galaxy S8 on Antutu's CPU bench? What does AnTuTu's UX score actually leasure? What is the methodology?

You have the tone of someone who understands this stuff, but you need to back it up with a display of knowledge at some point. Help me understand your point, because right now, I'm lost.
 

giga

Member
Lol how is optimizing an app to an architechure cheating? Thats basically apples whole advantage.

Optimizing an app to an architecture? It's literally raising thermal limits and CPU frequencies because it detected it was running a certain benchmark. Literally cheating because it's not what happens at any other time when using the phone.
 
Lol how is optimizing an app to an architechure cheating? Thats basically apples whole advantage.

You have it exactly backwards. This is an example of optimising an architecture to an app. Do you think Apple's whole advantage is optimising their SOC for synthetic benchmarks? Do you have any proof or documentation for that?
 
lol @ apple sheeple thinking their phone is the best. I’m with rambis, this is not remarkable or impressive, Samsung is the best maker of exploding phones out there.
 

Fhtagn

Member
Lol how is optimizing an app to an architechure cheating? Thats basically apples whole advantage.

That's not what's happening here.

https://www.anandtech.com/show/7384/state-of-cheating-in-android-benchmarks

1) On the Exynos 5410, Samsung was detecting the presence of certain benchmarks and raising thermal limits (and thus max GPU frequency) in order to gain an edge on those benchmarks, and

2) On both Snapdragon 600 and Exynos 5410 SGS4 platforms, Samsung was detecting the presence of certain benchmarks and automatically driving CPU voltage/frequency to their highest state right away. Also on Snapdragon platforms, all cores are plugged in immediately upon benchmark detect.

That's not optimizing, that's detecting that a benchmark is running and changing the state of the CPU to get performance you wouldn't get at any other time.

I would love to see an updated article on this to know how common these "optimizations" are in 2017, as this is from 2013.
 

Redmoon

Member
While it's nice to see a sector of computing tech advance performance wise this much whereas everything else seems stagnant, I'd personally I'd be all over an iPhone with the performance of my 6s or even a 7 but with a 3500mAh battery or more. Performance of my 6s is already more than what I need, but bat life on my heaviest usage(safari, source twitch streams) isnt that much last I timed it.

Doesnt help my 6s has also lost ~13% charge capacity :(
 

rambis

Banned
So now Geekbench is CPU dependant, but before you were arguing that resolution had a large effect on Geekbench scores. Which is it?

We discussed this, there are multiple tests. There are GPU tests in geekbench. Most of the workloads are CPU based. Idk where the contradiction is.



Why is Antutu better? It seems to be another synthetic benchmark, so why do you seem to refer to it as a real world benchmark? It doesn't even split out single core and multicore performance. Why is it better to obfuscate that info?


Why does a $350 iPhone SE with a two year old SOC pants a $750 Galaxy S8 on Antutu's CPU bench? What does AnTuTu's UX score actually leasure? What is the methodology?

There are more real world loads in Antutu.
http://www.makeuseof.com/tag/antutu-benchmark-measure/

Also I would say most CPU benchmarks give you one compiled score rather than splitting single/multicore.

You have the tone of someone who understands this stuff, but you need to back it up with a display of knowledge at some point. Help me understand your point, because right now, I'm lost.

You do seem lost but Im not about to spend my Saturday swapping essays with you.

Optimizing an app to an architecture? It's literally raising thermal limits and CPU frequencies because it detected it was running a certain benchmark. Literally cheating because it's not what happens at any other time when using the phone.

You have it exactly backwards. This is an example of optimising an architecture to an app. Do you think Apple's whole advantage is optimising their SOC for synthetic benchmarks? Do you have any proof or documentation for that?


Eh you are given a wide variety of performance enhancing modes and tools on the S8, particularly in developer mode. As long as the clocks are accurately reported I don't see this as cheating at all.

And platform specific optimizations are Apple's main advantage.
 

NeOak

Member
If only Apple didn't nerf the fucking modem.

Snapdragon X16 nerfed to shit to keep parity with a shit Intel modem.
 

Fhtagn

Member
Eh you are given a wide variety of performance enhancing modes and tools on the S8, particularly in developer mode. As long as the clocks are accurately reported I don't see this as cheating at all.

Again, that's not what happened. These devices would detect that a specific benchmark was running and then temporarily change the rules just for that benchmark, in ways you couldn't access otherwise.

If you can't acknowledge that as cheating, then there's no point in continuing this discussion.
 

rambis

Banned
It's pretty obvious that you don't understand any of this at the level necessary to explain it. Have a nice Saturday.

Lol thanks.

Again, that's not what happened. These devices would detect that a specific benchmark was running and then temporarily change the rules just for that benchmark, in ways you couldn't access otherwise.

If you can't acknowledge that as cheating, then there's no point in continuing this discussion.

You can, very easily. Again, as long as the clocks are reported to the app accurately I don't see the issue...
 

Fhtagn

Member
Lol thanks.



You can, very easily. Again, as long as the clocks are reported to the app accurately I don't see the issue...

You don't see the issue because you don't want to. Read the article linked above and pay attention to the fact that the clock doesn't change but the test is still being gamed.
 

ZOONAMI

Junior Member
If only Apple didn't nerf the fucking modem.

Snapdragon X16 nerfed to shit to keep parity with a shit Intel modem.

I dunno, I just got 20 down and 2 up with it on T-Mobile inside a huge art museum several floors above me, 3 bars on LTE. That’s pretty good imo given the situation. I’ll run another speed test outdoors later. People I work with here are surprised I even have service.

Edit: 40 down and 45 up outside. Considering that’s almost the same upload (less than half the download though) of my 100/50 fiber at home I’ll take it.

If a non nerfed x16 would do better that sucks I guess, but given the speed is good enough for me for virtually any usage scenario I can’t really complain. I also only have 6gb of lte data so I limit downloading a ton of stuff to WiFi.
 
Because everyone else in the mobile sphere is still years behind Apple, despite Apple hitting diminishing returns.

Yes, but what you fail to see is that none of this matters because Android is so good, it causes goalposts to spontaneously shift whenever discussing the performance of phones running that OS.
 

Spinluck

Member
The thing is, it will be still faster than next years high end android phone too. Apple has been destroying Android phones still 2 gens ahead. The 6S still beats the Note 8 in these tests.

Note 8 came slightly before, didn’t match it. Pixel 2 is coming out soon, probably won’t match it. It’ll be interesting to see how the S9 fares.

That's one way to simplify it. Too bad it's almost twice as fast as the nearest competition, whereas said competition could only manage a phone with about the same performance as Apple's year-old iPhone 7.

Gotta go fast
 

McLovin

Member
T mobile has this great deal where i can trade in my 6 plus for a $300 credit towards an 8. I was gonna switch to android but i think I'll give the iphone another shot. Pretty much 90% of the problems I had with my 6 plus was that I bought the 16gb model. Made me feel like the phone was a peace of useless junk whenever I really needed the space.
 

mrkgoo

Member
I got to see an iphone 8 in person ever yesterday. It actually looks really cool. I think my mind will explode when I see a X.
 

blu

Wants the largest console games publisher to avoid Nintendo's platforms.
See my bolded replies above.

I'm not missing this point. I've already acknowledged how impressive Apple's CPU is. My post was a reply to post made to post comment on comparisons to Intel CPUs. Some of the posters that replied to what I posted clearly didn't read the article because there replies we're basically, "see, the A11 is very fast!"
You didn't lie, you rushed into the thread with a 'lol, fanboys' strawman, using the very definition of a fanboy-fodder article as a 'I'm gonna show 'em' counter-argument. In reality none of the serious posters in this thread as much as suggested Apple should transplant their phone SKU into a macbook, evicting Intel. Everybody impressed with the single-core CPU performance of the A11 have been extrapolating what would happen if Apple used that uarch in a higher-end SKU - i.e. one with higher TDP, more cores, more cache, wider RAM buses, etc. Additionally, people have been speculating about the prospect to a productivity enclosure hosting a new-gen iphone. So get off the soapbox, and enjoy your stay if you have anything productive to add to the discussion.
 
Oh wow, cheating on benchmarks smh.

Cheating on benchmarks happens all the time in the smartphone sphere. There aren't really any actual good benchmarks for overall phone performance. Most of the ones which exist haven't been updated in forever and are generally easy to cheat on.

Because of recent advances in phone storage performance, many old phones have completely irrelevant benchmarks compared to new ones which can read and write to storage orders of magnitude faster. Just as with PCs and SSDs, nothing is more revolutionary to overall system performance than storage performance. Apple has a massive lead in storage performance over Android devices which is why apps load so much faster, plus this helps with video encoding or any task where you are streaming data to and from storage.

Saying the iP8 can encode video much faster than Androids is completely irrelevant because Apple has undoubtedly implemented their own acceleration which uses the GPU they designed themselves and the OS and software they coded themselves. Androids on the other hand are using off-the-shelf or semi-custom hardware designs and OS and it's hard to imagine anything is being properly accelerated or uses any kind of custom code. It's the difference between consoles and PC in terms of game optimization and we can now see how stark the difference is on smartphone platforms.
 
Cheating on benchmarks happens all the time in the smartphone sphere. There aren't really any actual good benchmarks for overall phone performance. Most of the ones which exist haven't been updated in forever and are generally easy to cheat on.

Because of recent advances in phone storage performance, many old phones have completely irrelevant benchmarks compared to new ones which can read and write to storage orders of magnitude faster. Just as with PCs and SSDs, nothing is more revolutionary to overall system performance than storage performance. Apple has a massive lead in storage performance over Android devices which is why apps load so much faster, plus this helps with video encoding or any task where you are streaming data to and from storage.

Saying the iP8 can encode video much faster than Androids is completely irrelevant because Apple has undoubtedly implemented their own acceleration which uses the GPU they designed themselves and the OS and software they coded themselves. Androids on the other hand are using off-the-shelf or semi-custom hardware designs and OS and it's hard to imagine anything is being properly accelerated or uses any kind of custom code. It's the difference between consoles and PC in terms of game optimization and we can now see how stark the difference is on smartphone platforms.

It's not irrelevant to people who encode video on their phones. The part that is irrelevant to them is how or why encoding is faster. They don't care if something is designed in house or off-the-shelf, all that matters is how it performs.
 
Cool, though I'm not sure I care, phones power stopped being all that relevant for my after the iPhone 4.

Makes me wonder how powerful the x will be tho.
 

Jigolo

Member
As much as I love my 7+, every day I find it harder and harder to not go crazy and buy an 8+ to replace it even though I got it in April.

But I can do it. I can do it. I don't need an 8+. I want it but I don't need it. I can hold off a year or two. Yeah. I can do that.
I've been feeling the same way about my 6S+. I'm trying to hold it till the 2019 iPhone though. Don't really care about specs that much myself but by then what can we expect? An A13 is given but with 6GB of RAM maybe? Well whatever, I'm looking more forward to the changes in the ios 13 or w/e it'll be called and iphone design changes (smaller or no notch pls)
 
Top Bottom