iPhone 4S benchmarked, beats the competion with ease

The iPhone 4S which was met with modest regard when unveiled is quickly getting some positive attention. First, the device sold 1 million units in 24 hours and now it has been benchmarked which shows the true power of the device.

Anandtech did the dirty work of putting the iPhone 4S to the test and if you thought the device would only compete with Android's best, think again. The iPhone 4S handily beat nearly every Android device out their, including tablets. The Samsung Galaxy S II and Droid Bionic were both handily beat by Apple's latest device which should put some perspective on how powerful this device actually is. The charts below are from AnandTech:

 

 

 

There are several more comparison charts over at AnandTech which are worth checking out as well. While it is clear that the iPhone 4S does beat the pack in nearly every category, there is a good possibility that Ice Cream Sandwich, the next major update to the Android platform, could help to level the playing field in terms of performance.

But, as it stands right now, the iPhone 4S beats the pack and is king of the benchmarks World which should put to rest some of the claims that Apple is only passing along a warmed over iPhone 4.

Apple devices will typically be on the market for up to a year without replacement. Knowing this, will it be able to fend off the next generation of Android devices in terms of performance or will the host of Android competitors be able to best Apple?

Thanks for the tip Vice 

Report a problem with article
Previous Story

Steve Jobs' Apple memorial to be held on Oct. 19

Next Story

Microsoft launches site to promote browser security

66 Comments

Commenting is disabled on this article.

Brad, Apple really does seem to be the leading innovator if they are beating out all other company's products! Apple has a great reputation, which may be why it's always dominating the market. Do you think there will ever be a chance for a different company, like Android, to pull in the lead?

Samsung Galaxy S2 running CyanogenMod 7 (Android 2.3.7) gives me 1700~ in the stock Android browser and 1500~ in Opera Mobile running Sunspider.

Just scored 3660ms in SunSpider on my HTC HD2 (yes, two years old) running Android 2.3.5 (MIUI) at 1Ghz. using the stock MIUI browser.

This phone will NEVER DIE.

I'm shocked that my trusty ol' HD2 (running Android 2.3.5) scored 2962 using Firefox. The browser generally feels slower than any other browsers I've tried on Android, but I guess they got SunSpider right.

I got ~5900 with the Android browser.

people need to know: benchmark does not mean anything, and pure hardware spec does not mean crap. android is slower than iphone is because: 1. android os is not optimized to use hardware graphic acceleration, and 2. most manufacturers just slap on bare minimum drivers on the devices. if you have an android os that properly supports all the graphics chip awesomeness on os level, properly optimized drivers, ext4 filesystem, on the current hardware, it will blow pass iphone on a lot of benchmarks

The sad part of this is every android phone in this is old Even the SGS2 is old compared to the iPhone 4s. Put the 4s up against the Nexus Prime and then lets see some numbers. New vs New. Not New vs Old by a year for specs (not OS)

Ok, specs are nice. Fair enough. Screen is crap, SAMOLED+ is better, no removable storage so you are stuck with 64GB and it runs iOS which is not going to make any use of all that power.

Flash android to it (hopefully some devs works that out) and then the specs will mean somthing.

People got bad experiances with android in its early days, remembering that most manufacturers dont update their devices most people are stuck on 1.5.

Running my galaxy 5 with MADTEAM Cyanogen makes it awesome. Triple the battery life too

Auzeras said,
Ok, specs are nice. Fair enough. Screen is crap, SAMOLED+ is better, no removable storage so you are stuck with 64GB and it runs iOS which is not going to make any use of all that power.

Flash android to it (hopefully some devs works that out) and then the specs will mean somthing.

You realise you just contradicted yourself? You slam the hardware features than suggest by simply changing the software the "specs will mean something"...so which is it? Is the hardware good or bad?

Kinda shows how far away mobile devices are from desktops. at least a 10x difference from the best mobile device listed to a couple year old computer i just tried it on.

wow... benchmarks and on phones to show.... numbers!
number that even if they are real mean nothing in real world usage.

the worse thing its when people care about these numbers and use these numbers like if they were written by gods.

but oh well...

EmilyTheStrange said,
wow... benchmarks and on phones to show.... numbers!
number that even if they are real mean nothing in real world usage.

the worse thing its when people care about these numbers and use these numbers like if they were written by gods.

but oh well...

Exactly.. the iPhone 4S might beat my phone in terms of numbers but my phone has a much bigger, brighter, better looking display for watching media/playing games on

No, that in fact that proves right all the review sites when talk about ip4 having the smoothest and unglitched operation amongst competition. I have Sam GIIS yeah bigger screen good for videow to the point my thumb can reach it, and definitely snappy but not without its glitches. Overall i liked a tad more the smoothness of Iphone. And Android store is lol..

And seems that 1 million people that in a single day preordered the 4s that might think the same.

Hardcore Til I Die said,

Exactly.. the iPhone 4S might beat my phone in terms of numbers but my phone has a much bigger, brighter, better looking display for watching media/playing games on

Hardcore Til I Die said,

Exactly.. the iPhone 4S might beat my phone in terms of numbers but my phone has a much bigger, brighter, better looking display for watching media/playing games on

Display test 50% brightness 100% brightness
Black, cd/m2 White, cd/m2 Contrast ratio Black, cd/m2 White, cd/m2 Contrast ratio
LG Optimus Black P970 0.27 332 1228 0.65 749 1161
Apple iPhone 4 0.14 189 1341 0.39 483 1242
HTC Sensation 0.21 173 809 0.61 438 720
Samsung I9000 Galaxy S 0 263 ∞ 0 395 ∞
Sony Ericsson XPERIA Arc 0.03 34 1078 0.33 394 1207
Samsung I9100 Galaxy S II 0 231 ∞ 0 362 ∞
BlackBerry Bold Touch 9900 0.29 403 1376 0.47 618 1304
BlackBerry Torch 9860 0.29 426 1456 0.47 665 1413

I assume you have an Optimus?

I have a Galaxy S2 and scored 3096ms on Sunspider javascript benchmark

Would be interesting to know what version of the firmware they were using as that's a pretty substantial difference.

It does seem somewhat unfair that they tested using Android 2.3.4 when Android 2.3.5 is out but there you go

Before people jump off the IE9/WP7/Mango sucks cliff...

1) Sunspider is DESIGNED for the WebKit browser. (This should be a 'clue'.)
2) IE9s optimizations are prevented from running on the test:
• As the previous controversy with theIE9 desktop version, this test
won't show real results as IE9 runs faster in the 'real world' than it
can just running mindless loops.
• IE9 is using prediction, and additional compiling techniques beyond the traditional
JIT technologies used by WebKit. This is where people need to read the whitepaper
and have some compiler understanding.

*PS People do know that the JIT in both Safari and Chrome (Nitro/V8) are based off Microsoft JIT designs, which were abandoned during the IE8 time-frame, right?

IE9 uses several new ways of thinking beyond just a JIT, and recursive tests like SunSpider can be instant in IE9, thus making the 'benchmark' seem to not be working, so this is not allowed when running SunSpider type tests. (Which is the opposite of most browsers, as they optimize to SunSpider, and Microsoft intentionally slows IE9/IE10 so it don't seem be cheating with prediction. Microsoft's Javascript tests are better for IE9/IE10 and comparisons.)


This goes back to the argument, whether a browser should be a mindless document display Application or a real- time web content compiling technology.

IE9 and IE10 treat all web content like Applications and thus compile and use prediction and other 'compiler' type optimizations on the content.

Webkit DOES NOT, and is just a good document/page display Application. (The JIT is one piece, imagine using more advanced JIT concepts and applying it all the pieces that are not Javascript, like network requests, HTML parsing, et.c)


So WP7 users don't freak, as you have seen, WP7 seems really fast at browsing, cause it is...

We have yet to test an Android based phone that can compete with any WP7 7.5 device for web browser - regular real world sites. Even on WP7 7.0 it was hard to find a dual core Android that could keep up in real world browsing, and WP7 used IE7/8 that was of the older browser model that worked like (Chrome/Safari/Firefox/Opera).

PS IE9 on WP7 is one area a dual core is a good argument, because as fast as WP7 is at single core multi-tasking and speed, the IE9 engine is designed around multiple cores/cpus, and is written for shoving the background compiler technologies to be running on a secondary core. So dual core WP7 devices with IE9 will really smoke. (Still wonder if Microsoft will wait til the NT WP8 kernel though instead of the WinCE 7.x multi-tasking core optimizations.)

Nice find Avenger . But on the desktop IE9 is blazingly fast on Sunspider like the WebKit browsers; If not faster. Why is it slower on mobile specifically ?
BTW could you post a link to the whitepaper

gawicks said,
Nice find Avenger . But on the desktop IE9 is blazingly fast on Sunspider like the WebKit browsers; If not faster. Why is it slower on mobile specifically ?
BTW could you post a link to the whitepaper

Good question, but I'm not convinced my phone was giving it all that it could. I ran the test 6 times, and my phone didn't even get hot

Looking forward to the battery life... My 3GS lasts a few hours before I have to charge it. I go everywhere with a battery pack just in case I need to charge.

jafoman said,
Looking forward to the battery life... My 3GS lasts a few hours before I have to charge it. I go everywhere with a battery pack just in case I need to charge.

just buy a second battery to have charged, just in case. oh, wait.....

Buttus said,

just buy a second battery to have charged, just in case. oh, wait.....

That's basically what a battery pack is anyway..oh wait..

DomZ said,

That's basically what a battery pack is anyway..oh wait..

yea, i'd love to have a battery sticking out of the bottom of my phone...

scorp508 said,
Not a single WP7 entry?

Nope, only Apples unreleased version vs Androids current release. Why not wait until Android gets their newest version out to rival ios5.

Gaffney said,

Nope, only Apples unreleased version vs Androids current release. Why not wait until Android gets their newest version out to rival ios5.
This would be a valid argument if there were a realistic chance of Ice Cream Sandwich coming out before tomorrow, since that's when iOS 5 comes out.

It will be even more interesting when tested with the HTC Titan with its 1.5Ghz processor.

SunSpider has always been a WebKit centric test, it would be a more complete test if Peacekeeper and Microsoft's IE Test Drive tests are run also.

Lexcyn said,
Google really needs to add hardware acceleration to their interface ... software does not cut it any more.

ICS will do this.

Renato456 said,
Compare to Nexus Prime and then see...

We will as soon as someone gets their hands on one don't worry

neufuse said,
any phones running windows phone 7 tested?

Probably, but they would be too scared to insert those results.

Coi said,

Probably, but they would be too scared to insert those results.

I got over 9000 (!) on my HTC surround

Which is actually kind of surprising being that JS is supposed to be accelerated on Mango, and speedtests usually run quite quickly

neufuse said,
any phones running windows phone 7 tested?

I just tested my LG Optimus 7 running WP 7.5. It scored 9732.6 ms +- 1%.
Not a stellar benchmark, but in everyday use it works just fine.

Sraf said,

I got over 9000 (!) on my HTC surround

Which is actually kind of surprising being that JS is supposed to be accelerated on Mango, and speedtests usually run quite quickly

I got 9700 ms on my HTC Trophy (Verizon). However, browsing the internet on this device is a great experience. It's just as good if not better than my previous Android phone running CM7 or stock. Granted there is no flash..

For the SunSpider test on my HTC Arrive (Mango) I got 9653. I was expecting better to be honest.

Not that I feel my phone is sluggish while browsing or anything I just don't like getting beat by apple.

Edited by Nimdock, Oct 11 2011, 3:40pm :

Coi said,

Probably, but they would be too scared to insert those results.

Doubtful - the older WP7 devices run around the 9000+ mark - be interesting to see the latest ones but even then probably not at the top.

Byron_Hinson said,
....

Considering the reference architecture hasn't changed, only to lower the minimum specs to allow for entry level, throw away smart phones, I would suspect even the newest phones will score in the same ballpark as existing WP7 phones.

Also, is there even OpenGL ES 2.0 on WP7 or would that be through some Dx wrapper? Since developers can't build native code on WP7 unless they have exclusive permission from MS to do so, I don't think we could find a comparison.

Second one is about graphics, doesn't say everything about the CPU >.>
Btw, wasn't iPad 2 clocked at 900 MHz?

Coi said,
Second one is about graphics, doesn't say everything about the CPU >.>
Btw, wasn't iPad 2 clocked at 900 MHz?

The bigger problem is that benchmarks are not indicative of real world gameplay. The Tegra 3 chips will out perform anything apple can throw at them. Also we know iPhone is the best selling phone, but Android is the best selling OS, causal game developers [the majority] would be best to develop for android first, since they would get more marketshare.

Nexus- said,

The bigger problem is that benchmarks are not indicative of real world gameplay. The Tegra 3 chips will out perform anything apple can throw at them.

The next chip that is out will out perform the current chips that are out. Also, the sky is blue. As far as "anything that apple can throw at them" what do you base this on? At some point Apple will have something faster than a Tegra 3. Thinking otherwise is just being ridiculous.

Also we know iPhone is the best selling phone, but Android is the best selling OS, causal game developers [the majority] would be best to develop for android first, since they would get more marketshare.

And deal with the fragmentation of the Android market? Developers want to maximize profit, and that doesn't always mean targeting the largest OS marketshare that means targeting the largest handheld marketshare. Right now iPhone is boss at that. Need evidence? Compare game selection on iOS App Store vs Android Market.

Nexus- said,
Android is the best selling OS, causal game developers [the majority] would be best to develop for android first, since they would get more marketshare.

No.
You need to rephrase, you imply that Android as an OS has a price, it does not.

Secondly, depending on what a developers objective is regarding mobile apps, Android may not be the best.

If you build ad-driven apps, Windows Phone 7 has the highest click-thru revenue generation of all smartphones. So, if you make money from ads, you're better off building a windows phone app, and worry about the ad-platforms from the other OS providers later.