Tegra 3 versus the Galaxy Nexus in a benchmark.

[sushiglobster]

Yea I know, weird title for the thread, but the link shows the new Asus Transformer Prime being compared to the Galaxy Nexus in a benchmark.

The Galaxy Nexus also scores slightly higher than the Galaxy S2.

I have no idea how legit the benchmark is, but anyways...take a gander here: Transformer Prime gets mysteriously and imperiously benchmarked

 

[Adrynalyne]

That should put to rest some of the concerns about the gpu. Imagine what the score would have been if the GN was the same resolution as the GS II.

 

[ManBearPig618]

Yet the Cult of Exynos will still flood this thread to explain why the OMAP was obsolete in 1995.

 

[Eric Kane]

It should put it to rest, but it won't. They'll start complaining about how the Nexus isn't in the same class as the Asus Prime and how "Nah dude, I'm waiting for quad core now."

 

[daitsuwamono]

Before anyone gets any ideas, the Transformer Prime is a tablet, not a phone. No Tegra 3 phones have been announced, though they may start appearing in February.

..and hopefully I'll be wealthy enough by then to get in on some of that sick quad-core action.

 

[Adrynalyne]

LOL.

News flash:



BREAKING:
Single core GPU keeps up with quad core GPU. Riots everywhere.

More at 5.

 

[Eric Kane]

Can someone develop an app for the facial recognition software on the Nexus that scans for dumbasses? And then even if they set their face to be the one who unlocks it, it won't unlock and thus becomes useless?

We don't pick our phones...our phones pick us.

 

[CynicX]

I've liked the tegra 2 its great with battery more then fast enough for my needs, over clocks great. Only time I've seen it start to choke up is using a PlayStation one emulator with a LOT going on on the screen.

Is it possible to buy a phone or tablet and say.....man the CPU sucks!!! Lol

 

[daitsuwamono]

BREAKING:
Single core GPU keeps up with quad core GPU. Riots everywhere.

More at 5.

You can see that CPU integer and floating-point performance are approximately double that of the dual-core Galaxy Nexus. The GPU in the Transformer Prime is supposedly 12-core, but it's apparent that this benchmarking utility wasn't designed to utilise multiple hardware threads in the GPU, instead using just one as it does with the GN (which has only one GPU core). While it may not be completely reasonable to expect a full twelve-fold increase over a single-core GPU at a comparable clock rate, it'll be better than the one-to-one parity with the Galaxy Nexus's single GPU that this benchmark reports.

 

[Adrynalyne]

Pay attention to what I was referring to.


Hint: It wasn't the Tablet.

 

[anon(512898)]

You can see that CPU integer and floating-point performance are approximately double that of the dual-core Galaxy Nexus. The GPU in the Transformer Prime is supposedly 12-core, but it's apparent that this benchmarking utility wasn't designed to utilise multiple hardware threads in the GPU, instead using just one as it does with the GN (which has only one GPU core). While it may not be completely reasonable to expect a full twelve-fold increase over a single-core GPU at a comparable clock rate, it'll be better than the one-to-one parity with the Galaxy Nexus's single GPU that this benchmark reports.

actually the reason all the gpu scores are the same is because they all hit they 30fps frame rate cap that is present in android while rendering the benchmark scenes

 

[Adrynalyne]

Nah. Quadrant, and Nenamark hit over that regularly.

 

[anon(512898)]

Nah. Quadrant, and Nenamark hit over that regularly.

I thought quadrant was still affected by that, I guess there's nothing to see here then folks

 

[Adrynalyne]

I thought quadrant was still affected by that, I guess there's nothing to see here then folks

I think that was an artificial limit imposed by HTC for HDMI on the Evo4G. I never saw it happen anywhere else anyway.

 

[whiteshadow001]

It beat the s2 so that must be a good sign especially since it has a better screen and that usually messes with benchmarks

Sent from my Thunderbolt using tapatalk

 

[Adrynalyne]

A lot of people never took into consideration that the GPU wasn't just clocked higher, but fed by dual channel ram. That is not the case with the Exynos, I do not think.

 

[kca2000]

Nah. Quadrant, and Nenamark hit over that regularly.

Speaking of quadrant, I saw on another forum that the G-nex runs a 1700. I never knew quadrant varied THAT much. Thats half of what a gs2 is capable of putting up yet this test ranks them as the nexus having a slight edge overall. NOT COMPLAINING, just saying, Quadrant seems to be referred to alot in reviews i read and thats a serious difference vs what were seeing here on Antutu.

EDIT: sorry, heres the link to the video. http://www.droidforums.net/forum/galaxy-nexus/178069-galaxy-nexus-gets-quadrant-benchmarked.html

 

[Droid800]

Speaking of quadrant, I saw on another forum that the G-nex runs a 1700. I never knew quadrant varied THAT much. Thats half of what a gs2 is capable of putting up yet this test ranks them as the nexus having a slight edge overall. NOT COMPLAINING, just saying, Quadrant seems to be referred to alot in reviews i read and thats a serious difference vs what were seeing here on Antutu.

EDIT: sorry, heres the link to the video. http://www.droidforums.net/forum/galaxy-nexus/178069-galaxy-nexus-gets-quadrant-benchmarked.html

It hasn't been released yet, so quadrant hasn't been updated for it. (yes, it matters)

That, and the screen resolution mean quadrant is useless because it weights resolution really heavily.

 

[kca2000]

It hasn't been released yet, so quadrant hasn't been updated for it. (yes, it matters)

That, and the screen resolution mean quadrant is useless because it weights resolution really heavily.

Thanks for explaining that. I was not aware the software had to be updated to test the phone properly. I am happy however that the Nexus scored right up with the Gs2 with a proper test since thats the slickest/smoothest phone ive used yet.