Benchmarking Questions

[Murph5150]

My GNex is rooted and overclocked to 1.6 GHz. Why is it in benchmark tests the HTC X and S3 still score way higher? Are their processors that much better?

 

[oxymoron]

Short answer = yes.

Dual core vs quad core. The GS3 also has twice the RAM.

*clever signature*

 

[2defmouze]

Benchmarks are generally completely worhtless, meaningless numbers..lol. Its a complete waste of time to run them, if you ask me. Its a known fact that developers can produce ROMs/kernels that score high in benchmarks but perform sluggishly, and the other way around as well. You can run the same benchmark 3 times and you'll get different scores each time, it depends on so many random factors from the apps you have installed to how hot your device is.

Best thing IMO is to not bother with them. Just judge performance based on how it feels to you. Life will be so much less complicated

 

[Murph5150]

Benchmarks are generally completely worhtless, meaningless numbers..lol. Its a complete waste of time to run them, if you ask me. Its a known fact that developers can produce ROMs/kernels that score high in benchmarks but perform sluggishly, and the other way around as well. You can run the same benchmark 3 times and you'll get different scores each time, it depends on so many random factors from the apps you have installed to how hot your device is.

Best thing IMO is to not bother with them. Just judge performance based on how it feels to you. Life will be so much less complicated

I hear ya. Still would be nice to know, like my Camaro when it's dyno-tested, has the power under the hood. I didn't anticipate still getting "spanked" in benchmarks after rooting. But the GNex does run nicely, I must say

 

[oxymoron]

Yeah, mine benchmarks lower on Jelly Bean with a higher clock than ICS with a lower clock... but in the real world, it's much faster on JB.

Probably because a lot of tweaks were made to the way the GPU is handled, triple buffering, etc. The end result is a much smoother device, and that has nothing to do with benchmarks.

*clever signature*

 

[dmmarck]

Yeah, mine benchmarks lower on Jelly Bean with a higher clock than ICS with a lower clock... but in the real world, it's much faster on JB.

Probably because a lot of tweaks were made to the way the GPU is handled, triple buffering, etc. The end result is a much smoother device, and that has nothing to do with benchmarks.

*clever signature*

All of this . I have yet to approach my best quadrant on ICS with JB, but JB is just hands down quicker. Quadrants are great to pass the time and neat to test kernels with, but in reality get useless fairly quickly.