Six Optimus V ROM benchmarks from Linpack and Quandrant

[Hugh Johnson]

We have a few different choice for ROMs on our rooted Optimus V phones so I decided to see if one is significantly better than the others. To skip to the summary, no you won't really perceive any difference in speeds if the scores are any indication.

I tested a stock 2.2 nandroid backup, Stock Virgin Mobile + Xionia Kernel (BETA 2), RodimusROM v1.2.3 + xionia , Aphrodite 1.3 (Aphrodite 2 beta was released the following day after testing), aospCMod | AOSP 2.3.3 / CM7 (3/25/11), and Zefie's CM7 (03/01/2011).

I tried to keep the testing as basic and uniform as possible therefore the phone wasn't overclocked at all. I installed gapps-gb-20110307-signed.zip on each ROM and Linpack and Quadrant through the ADB terminal. The only thing that I messed up was with forgetting to install the updated Google Apps on the stock 2.2, but it doesn't really matter because it was the lowest scoring anyway.

I ran each Linpack test twice and capture the second score. Scores will get slightly better after each run (possibly due to caching). Quadrant doesn't have a remotely decent benchmarking method - it just averages all the subscores together.

If you can't already tell that means a phone that runs like crap in the real world but has a fast processor may score higher than a mid-level phone like our Optimus V. Compare the scores from the 2.2 to 2.3 updates too. I never noticed that much of a performance increase between them.

Stock 2.2 nandroid backup

Stock Virgin Mobile + Xionia Kernel (BETA 2)

RodimusROM v1.2.3 + xionia

Aphoridte 1.3

aospCMod | AOSP 2.3.3 / CM7 (3/25/11)

Zefie's CM7 (03/01/2011)

So it really just comes down to what you want in your ROM. If you want 2.2 with just your choice of apps Aphrodite is the way to go. If you want it with a little more meat use the Rodimus ROM. If you want 2.3 you're going to have to use a CM7 port so give aospCM or Zefie's CM7 a shot.

PS. Zefie's has a built in file manger so installing APKs is easy. That's all that makes it my preferred CM7.

 

[nickmiller71]

That's a great comparison! Seems quite fair, though I'm really really really surprised that Zefie's CM7 scored a 1062 in Quandrant, even though it hasn't been updated in quite some time...

 

[JerryScript]

I assume there was no overclocking used.

I've noticed that Linpack scores increase with overclocking.

Quadrant's scoring doesn't seem reliable enough to use for benchmark testing.

On Linpack, I get the same 7.5 avg that your results show. Overclocking using aospCM7 pushes the Linpack score up to 9.2-9.8.

 

[LeslieAnn]

My thanks to Hugh Johnson for this, it's interesting if nothing else.


Please keep in mind that these are synthetic benchmarks.
At least one of the 2.3 roms uses overclocking. Probably both.
Also, there is also some questions that have been raised about their benchmarks scores. It could be that the tests are beneficial to 2.3 roms vs. 2.2

While the CM7 roms look way faster, they don't necessarily feel that much different in use compared to the 2.2 roms. By the same token however, I can tell a difference between stock, and Stock + Xionia in terms of speed, it's just not much, but at the same time, I can't tell any difference between Stock + Xionia and Aphrodite. Which based the results above, you would think I could. Then again, since moving off the stock rom I have messed around with overclocking, so my perceptions may be off. Another oddity, is that all of the 2.2 roms use the same kernel and foundations, the only thing I can think is happening is that apps in memory are effecting the scores.


My advice is to run the rom that works best for you.



Also, I'm kind of glad Aphrodite 2.0 isn't included, I'm not sure its really ready for this yet.

 

[idbl_fanatic]

I just got done doing 5 runs on the stock vm + xionia and got an average of:

825 mhz
9.6156 linpack
589.4 quadrant

806 mhz
9.5934 linpack
597.2 quadrant


Sent from my LG Optimus V

 

[Hugh Johnson]

I don't think it's that surprising that Zefie's CM7 scored about 100 points higher than aospCM even though it's a little older. When you compare the 2.3 Gingerbread values to the previous 2.2 Froyo values you can see 100 points becomes a relatively small margin. I would really like to find out what 2.3 has in it that makes such an impact on Quadrant's CPU test.