3DMark Ice Storm Mobile Benchmark: Post your results

[Raduque]

Guess it's not low. What Rom you running?

I originally tried Perseus but switched to Jedi's kernels and Roms a few months ago and haven't looked back. (Sons of Android/SOAP or JediXP12/SaberKernel)

As for your question, 16GB Phone Memory + 16GB Micro SD Card. Its just easier to say 32GB. Had a 64GB Sandisk just crap out on me so stuck with 16GB for now.

Completely stock AT&T LK7, only difference is Perseus kernel, and I run it because it has "Privacy" mode.

As for the way you describe your storage, imo, it's misleading. You don't have a "32gb Note II", you have a 16gb Note 2 with a 16gb uSD card. By your logic, I can call mine a 48gb Note II. Sorry, it's just one of my pet peeves. It's like saying the CPU is 6.4ghz.

 

[thecapsaicinkid]

The bargain basement N4 outperforming everything again. Note 2 scores are lousy.

 

[dguy6789]

Motorola Droid Razr Maxx HD
Android 4.1.2

Ice Storm Score 6153
Graphics Score 6561
Physics Score 5052
Graphics Test 1 30.6fps
Graphics Test 2 26.7fps
Physics Test 16fps

What's interesting to me is that despite the same SOC this phone outscores the Galaxy S3 in every test and in the physics test by a huge margin. The Razr HD has always outscored the GS3 historically by a little in pretty much all benchmarks but it outscoring the GS3 physics score(using the Anandtech number) of 2802 with a score of 5052 is very curious. Could Touchwiz be to blame?

 

[ChronoReverse]

Not sure about Touchwiz but it could be something weird with the stock kernel since if you look at the physics score, it's obvious that only one core of the GS3 activated for Anand's test.

For the GPU part, the Razr HD might just have its Adreno 225 clocked higher. My One XL with the same specs as the GS3 scores basically the same except the physics test where it doubles the Anand result.

 

[jacktesterson]

Completely stock AT&T LK7, only difference is Perseus kernel, and I run it because it has "Privacy" mode.

As for the way you describe your storage, imo, it's misleading. You don't have a "32gb Note II", you have a 16gb Note 2 with a 16gb uSD card. By your logic, I can call mine a 48gb Note II. Sorry, it's just one of my pet peeves. It's like saying the CPU is 6.4ghz.

haha. You have too much to worry about I had it that way due to Sig limitations

Just for the record, 32GB Note II's are available if you buy the N7100

 

[swanysto]

Surprised not to see any GS3's yet. Here is mine(completely stock).

Ice Storm Score 4923
Graphics Score 6434
Physics Score 2702
Graphics Test 1 29.4fps
Graphics Test 2 26.7fps
Physics Test 8.6fps

Looks like my phone sucks.

 

[ChronoReverse]

Huh, the physics score on your phone has the same problem as Anand's. It looks like one of the cores isn't engaging (it's literally getting half the score of my otherwise identical One X).

I'm going to see if my friend's non-stock GS3 has this problem if it's using a different kernel governor.

 

[Fire&Blood]

Extreme setting:

Standard:

 

[lopri]

Ran it one more time before uninstalling it. Non-active tasks that sits in the memory don't seem to affect the scores very much. Actually the scores went up the 2nd time. (But the CPU scores are slightly down)

Still, my N4's CPU scores are unusually high compared to others? (I'm not bragging for there is nothing I have done to that effect. Just curious)

 

[ChronoReverse]

Are you stock?
My dual core S4 @ 1.5GHz got 5750ish which doubles to 11500. That would mean your phone is getting the expected score and everyone else is scoring _low_.

 

[lopri]

Yeah I'm stock. Maybe you're getting high, too. Mine is apparently high, not by my account, but by Futumark's standard (whatever that may be). You can see the gray scores, which are supposedly "expected" or "average" scores for the Nexus 4, next to my scores.

Per usual 3DMarks are useless for all intents and purposes.

 

[ChronoReverse]

I'm not the only one with the same hardware getting a similar score though. A lot of people also seems to be running into some bug causing the physics score in particular to score too low, as if a core failed to spool up.

 

[jacktesterson]

Scores are actually a little higher with my Note II with AOSP 4.2.2 based ROM

Note II
Stock clocks
Sons of Android 4 JB 4.2.2
SOAPKernel 5

 

[pakotlar]

Scores are actually a little higher with my Note II with AOSP 4.2.2 based ROM

Note II
Stock clocks
Sons of Android 4 JB 4.2.2
SOAPKernel 5

Those scores are incredible considering your Mali is @ 533mhz. Mine at 800mhz scores only 12fps on Gtest1, 25fps on Gtest2, and 25fps on the demo. I see that SoapKernel recently updated their Mali drivers, hopefully this comes over to my kernel soon (Perseus).

 

[thecapsaicinkid]

http://www.youtube.com/watch?v=5Q_AJe0ZI-E

n4 outperforming the One.

 

[pakotlar]

Note 2 @ 1800 CPU / 800 mhz GPU update (pegasus governor) 4.1.2:

Extreme: 3340
GPU Test 1 : 10.1
GPU Test 2 : 15.2

This is actually pretty impressive. Nexus 10 scores only 18.2fps in Test 2, and even the One achieves only a marginally better result 22.3fps (also 4.2.2 or newer mali drivers may provide significantly faster performance, for Note 2 as well judging by AOSP Note 2 results above).

Since GPU Test 1 is geometry heavy, it makes sense that the Mali 400 struggles (but still provides ~ 50% faster performance than HTC One X+). But it seems that its pixel shaders/fill rate actually competes favorably with newer devices, even @ 1080p. Of course it only needs to render 720P, so on balance it may be providing a GUI performance ~ HTC One levels (In fact @ 800mhz I can achieve a 1.2Gpixel fill rate in GLBench 2.7 @ 720P, and ~1Gpixel @ 1080P; HTC One scores ~1Gpixel @ 1080P).

Also, unlike most (not all) overclocked devices, I can actually run my Note 2 at these clocks and achieve > 24 hours battery life, with wifi or 4g always on, and ~6-7 hours screen time @ 75-100% brightness. My battery lasts far longer than my wife's iPhone5 on 4g/screen-on, id say 50% longer. My CPU/GPU is undevolted.

Anyway, since I don't really game on this device, but am looking at exemplary GUI performance, Note 2 impresses me.

Caveats: I don't know exactly what GUI performance depends on; I know fill-rate and blend rate are important however, and Mali400 seems to excel at least in fill-rate.

 

[ChronoReverse]

Since GPU Test 1 is geometry heavy, it makes sense that the Mali 400 struggles (but still provides ~ 50% faster performance than HTC One X+)

One thing to note is that the One X+ is a Tegra3 GPU which is just plain bad compared to its contemporaries. The X+ ostensibly replaces the X but the Adreno 225 in the X (ATT version) wipes the floor with it (the Tegra3 does have higher clocked quad A9's to take a larger lead in the CPU department).

 

[Kenmitch]

My AT&T Optimus G got the jelly bean update the other day!

Well I think this app is still a little buggy as my high score results for the standard run won't update for some reason or another. If I get a new high score for the extreme it'll update.

Seems like the phone throttles with heat which effects the scores. These are the best runs I've gotten so far....Very cool phone required to get anything remotely close to these results.

100% bone stock

I don't even game on my phone.

 

[pakotlar]

One thing to note is that the One X+ is a Tegra3 GPU which is just plain bad compared to its contemporaries. The X+ ostensibly replaces the X but the Adreno 225 in the X (ATT version) wipes the floor with it (the Tegra3 does have higher clocked quad A9's to take a larger lead in the CPU department).

Agreed.