Evo 4g/Evo 3D

[gTen]

Originally Posted by BlackDynamite

lol, like I said, you're a real piece of work. I point out that the Sensation finished significantly faster in a performance test, yet still scored lower. You, after arguing that the test is not valid, continue to argue why this test IS valid, and make a bunch of excuses for why this happened, and then say i'm the one making excuses. The bottom line is- the Sensation finished the quadrant bech test significantly faster than the Galaxy S 2. Period. Now you trying to justify that is not ME making excuses, but rather it is YOU making excuses as to WHY it happened.

I think you are mistaken...I do not need to justify anything or give excuses for anything..Quadrant clearly gave SGS2 a higher score PERIOD..you can say whatever you want but according to Quadrant the SGS 2 is faster then the Evo 3d...

The only thing I was trying to explain to you was how quadrant works and why the score is what it is..for YOUR educational purpose. The fact of the matter is SGS 2 won in Quadrant..if you do not wish to hear why thats your choice.

I know it is different. Never said otherwise. Doesn't change a thing. The bottom line is if you are testing the processing power on a device, it isn't a fair test if one is emulating a different resolution.

they arent emulating..they are rendering at the resolution set to them..if you set them to render at same resolution they will produce accurate benchmarks.

You are the one denying the facts here. But carry on making up your own world to live in. I'm done wasting time on you.

Yes, in my world 3000 > 1600 I do not know why thats different in yours :/

[BlackDynamite]

Originally Posted by gTen

Yes, in my world 3000 > 1600 I do not know why thats different in yours :/

You go through post after post saying quadrant is not valid. I point out that I was the one who first said it wasn't valid. So now you have several posts explaining why it actually IS valid. You then follow that up with a little jab implying that I am the one living in a fantasy world.

lol, like I said, you're a real piece of work. Carry on.

[gTen]

Originally Posted by BlackDynamite

You go through post after post saying quadrant is not valid. I point out that I was the one who first said it wasn't valid. So now you have several posts explaining why it actually IS valid. You then follow that up with a little jab implying that I am the one living in a fantasy world.

lol, like I said, you're a real piece of work. Carry on.

I never once said anything of the matter..let me repeat in English in 1 sentence:

"According to quadrant, SGS2 has more processing power then the Evo 3d."

Where does it say that Quadrant is valid? I just explained to you the results quadrant gave >.> lol...I tried to go into detail on why Quadrant gave which score and explain to you why one finished before the other but you did not wish to hear it so I just stated to you the facts in that 1 line.

The only benchmarks right now that are valid for new phones are:

#1) Smartbench and #2) GL Benchmark

as they adapt for new technologies better then the rest. I welcome you to do some research on them if you wish as you do not even want to listen to what I have to say.

And lastly, I did not jab you..I just dodged your jab and have you jab yourself

[BlackDynamite]

Originally Posted by gTen

I never once said anything of the matter..let me repeat in English in 1 sentence:

"According to quadrant, SGS2 has more processing power then the Evo 3d."

Where does it say that Quadrant is valid? I just explained to you the results quadrant gave >.> lol...I tried to go into detail on why Quadrant gave which score and explain to you why one finished before the other but you did not wish to hear it so I just stated to you the facts in that 1 line.

The only benchmarks right now that are valid for new phones are:

#1) Smartbench and #2) GL Benchmark

as they adapt for new technologies better then the rest. I welcome you to do some research on them if you wish as you do not even want to listen to what I have to say.

And lastly, I did not jab you..I just dodged your jab and have you jab yourself

And I will finish off by saying that I don't think any of those bench tests are valid. You yourself admitted that none of them are perfect, so they are all flawed in your opinion too. But I will say this- linpack is the most valid one of the bunch in my opinion, and it gave the better score to the Sensation. You can choose to say you prefer smartbench, but I don;t care. It doesn't mean I have to prefer it just because you do. Like I said, at least linpack is running at the native resolution on each device.

Now, back to your fantasy world.

[gTen]

Originally Posted by BlackDynamite

And I will finish off by saying that I don't think any of those bench tests are valid. You yourself admitted that none of them are perfect, so they are all flawed in your opinion too. But I will say this- linpack is the most valid one of the bunch in my opinion, and it gave the better score to the Sensation. You can choose to say you prefer smartbench, but I don;t care. It doesn't mean I have to prefer it just because you do. Like I said, at least linpack is running at the native resolution on each device.

Now, back to your fantasy world.

Linpack isn't running on any resolution...it only tests floating point...this is equivalent to Quadran'ts CPU score and GL benchmark's "CPU Performance: Float" benchmark...

aka its a benchmark that only measures 1 thing, the CPU and only in Float. And the score difference was 1%.

You can prefer it if you wish, but the benchmark fails to test the 2d and 3d GPU, CPU Int, I/O and memory.

BTW here is an official statement from the creator of linpack for android:

This test is more a reflection of the state of the Android Dalvik Virtual Machine than of the floating point performance of the underlying processor. Software written for an Android device is written using Java code that the Dalvik VM interprets at run time.

https://market.android.com/details?i...=search_result

Hence why I said Linpack only works for benching the same phone...so the author themselves said its not a good benchmark for cpu processors..now what?

There is no perfect benchmark yes, but Smartbench at the very least accounts for an overall performance of the device and is optimized to account for resolutions and dual cores.

[BlackDynamite]

Originally Posted by gTen

Linpack isn't running on any resolution...it only tests floating point...this is equivalent to Quadran'ts CPU score and GL benchmark's "CPU Performance: Float" benchmark...

aka its a benchmark that only measures 1 thing, the CPU and only in Float. And the score difference was 1%.

You can prefer it if you wish, but the benchmark fails to test the 2d and 3d GPU, CPU Int, I/O and memory.

BTW here is an official statement from the creator of linpack for android:



https://market.android.com/details?i...=search_result

Hence why I said Linpack only works for benching the same phone...so the author themselves said its not a good benchmark for cpu processors..now what?

There is no perfect benchmark yes, but Smartbench at the very least accounts for an overall performance of the device and is optimized to account for resolutions and dual cores.

He is not saying that THIS test doesn't work for testing CPU, he is saying that NO test works for testing CPU. He is clearly saying that since Android runs a Dalvik Java Virtual Machine, the best any test can do is test that virtual machine, rather than the actual machine. And that is also what I said- remember when I said that they weren't testing the dual core until the OS was designed that way?

The only way around this is with root access. Otherwise, what the linpack creator said is 100% correct- you are only testing the virtual machine and not the actual machine.

[gTen]

Originally Posted by BlackDynamite

He is not saying that THIS test doesn't work for testing CPU, he is saying that NO test works for testing CPU. He is clearly saying that since Android runs a Dalvik Java Virtual Machine, the best any test can do is test that virtual machine, rather than the actual machine. And that is also what I said- remember when I said that they weren't testing the dual core until the OS was designed that way?

The only way around this is with root access. Otherwise, what the linpack creator said is 100% correct- you are only testing the virtual machine and not the actual machine.

umm...

This test is more a reflection of the state of the Android Dalvik Virtual Machine than of the floating point performance of the underlying processor. Software written for an Android device is written using Java code that the Dalvik VM interprets at run time.

he is clearly saying "this test"....why do you put words in his mouth?

Now as for smartbench:

Smartbench 2011 is a multi-core friendly benchmark application that measures the overall performance of your smartphone. It reports both Productivity and Games Index to suit both productivity users and 3D gaming users.

Smartbench 2011 - Android app on AppBrain

[BlackDynamite]

Originally Posted by gTen

umm...



he is clearly saying "this test"....why do you put words in his mouth?

Now as for smartbench:



Smartbench 2011 - Android app on AppBrain

So wait- are you implying that smartbench is not running on the Dalvik Virtual Machine? Please clarify. Because either A: it IS running on the Virtusal Machine and therefore his statement would also apply to smartbench, or B, it is not running on the Virtual Machine so therefore his statement would not apply to smartbench.


If you agree that it IS running on the virtual machine, so therefore his statement also applies to smartbench, then I have to ask why the heck you are posting all this crap.

EDIT:
From his same description that you quoted (the next sentence after your quote):
"Software written for an Android device is written using Java code that the Dalvik VM interprets at run time."

Clearly, he is saying Android software runs on a virtual machine (which is why the benchmark tests the virtual machine and not the actual machine), and NOT just this app.

[gTen]

Originally Posted by BlackDynamite

So wait- are you implying that smartbench is not running on the Dalvik Virtual Machine? Please clarify. Because either A: it IS running on the Virtusal Machine and therefore his statement would also apply to smartbench, or B, it is not running on the Virtual Machine so therefore his statement would not apply to smartbench.


If you agree that it IS running on the virtual machine, so therefore his statement also applies to smartbench, then I have to ask why the heck you are posting all this crap.

EDIT:
From his same description that you quoted (the next sentence after your quote):
"Software written for an Android device is written using Java code that the Dalvik VM interprets at run time."

Clearly, he is saying Android software runs on a virtual machine (which is why the benchmark tests the virtual machine and not the actual machine), and NOT just this app.

Read up on SMP