Attend XDA's Second Annual Developer Conference, XDA:DevCon 2014!
5,743,150 Members 44,093 Now Online
XDA Developers Android and Mobile Development Forum

[Q] kitKat Performance

Tip us?
 
Journyman16
Old
(Last edited by Journyman16; 6th March 2014 at 04:13 AM.) Reason: Ask about battery
#1  
Journyman16's Avatar
Senior Member - OP
Thanks Meter 208
Posts: 486
Join Date: Aug 2012
Location: Bendigo

 
DONATE TO ME
Default [Q] kitKat Performance

Thought I'd move the questions to the top for ease of reading...

1. What I would like to know is how KK performance stacks up against 4.3?

2. How many KK users have benchmarked their phone before and after?

3. Does a KK phone get close to the Samsung Note 3 standard rating in Antutu?

4. How is battery life?

I've been very pleased with my Note 3 n9005. Even on stock the battery was good and performance was quick. I stuck with it for some time while I read up about Knox and other issues to do with rooting. I wanted to have the phone for a while to make sure I was well into the bathtub curve before I blew my warranty.

Mind you it might still be good anyway with Kogan as it's a grey import and Kogan have always been good with warranty in my experience.

Using RDLV I rooted my phone and did things like install TBup, Greenify and other root-based apps I like..I tried a variety of ROms - I had a Hong Ko0ng 32GB version so had to learn to fix PIT issues so I could flash 'normal' ROM's and not just ZGY ones - at each step I spent a while on that set up to make sure nothing was going wrong, but somewhere in my playing around I managed to trigger Knox - not sure how as it happened when I was simply repeating a build I had done before. C'est la vie! Now I could look at custom recoveries and ROM's.

I flashed Philz, made a nandroid and downloaded Omega v6.

I had used Omega on my S3 and was impressed so I got Omega v6 (4.3-based) as I wasn't ready to jump irrevocably into KK. There were too many issues being reported or questioned so I figured I'd wait a bit.

On stock I had run the Antutu benchmark and gotten above 35000 so I was a bit disappointed when with Omega the figure dropped to about 27000. It wasn't obvious in most things but playing video or games was a bit slower. I tried Faux kernel to see if it would help and got 22500 on Antutu.

I DL'd Alliance ROM, mainly because the Dev is still updating it where Omega 4.3 development has ceased. Very pleased with it and flashing Lean kernel took me back up to 35,400 on Antutu.

So, I am about ready to make the jump to KK - many of the initial issues are sorted out (so far - Samsung could always mess things up) and there are a lot of options for ROMs and jernels coming out.

I'll be doing some browsing on that but thought I might as well ask here. With Lean kernel and Alliance ROM and light use I am getting around 15 hours per 10% - a busier day gives me about 10 - 11 hours per 10%.

And that's with Wifi on all the time, BT on most of the time, GPS probably 50% - I think Greenify helps a lot as does debloat the Samsung stuff, but it is still damn good.

I'd like to NOT lose that kind of performance.
Never doubt there is Truth, just doubt that you have it!
 
Journyman16
Old
#2  
Journyman16's Avatar
Senior Member - OP
Thanks Meter 208
Posts: 486
Join Date: Aug 2012
Location: Bendigo

 
DONATE TO ME
Bump - still interested in KK experiences versus JB ROMs.

From my reading I am still thinking to stay with 4.3 - still a lot of questions around about KK.
Never doubt there is Truth, just doubt that you have it!
 
Rosli59564
Old
#3  
Senior Member
Thanks Meter 142
Posts: 388
Join Date: Mar 2013
Quote:
Originally Posted by Journyman16 View Post
Bump - still interested in KK experiences versus JB ROMs.

From my reading I am still thinking to stay with 4.3 - still a lot of questions around about KK.
Personally i don't see a better benchmark with KK,about the same with JB. Battery life however very much better than JB. On any custom rom I can only get about 4hrs screen time while with KK about 7hrs. That's a huge difference
 
Journyman16
Old
#4  
Journyman16's Avatar
Senior Member - OP
Thanks Meter 208
Posts: 486
Join Date: Aug 2012
Location: Bendigo

 
DONATE TO ME
Quote:
Originally Posted by Rosli59564 View Post
Personally i don't see a better benchmark with KK,about the same with JB. Battery life however very much better than JB. On any custom rom I can only get about 4hrs screen time while with KK about 7hrs. That's a huge difference
That's a big difference alright. I don't use a lot of screen time as mostly a light user, but I am getting between 3 and 5 days before charging is needed. But last reading was 22+hours for 85% with 1.5 hours screen on time.

I'm curious as to performance because of the articles about Samsung modding JB to work better just for benchmarks - apparently after being called on it they removed the mods for KK.
Never doubt there is Truth, just doubt that you have it!
 
pete4k
Old
#5  
Senior Member
Thanks Meter 55
Posts: 326
Join Date: Jan 2014
Quote:
Originally Posted by Journyman16 View Post

I'm curious as to performance because of the articles about Samsung modding JB to work better just for benchmarks - apparently after being called on it they removed the mods for KK.
As far as I know they ramp up frequency to maximum when the start of benchmark was detected. You can do same thing by yourself, if rooted: For example in ROM toolbox move minimum frequency slider to max, or switch governor to performance, just make sure you switch it back afterwords.
Samsung is stupid, instead of removing the mod, they should have made it user adjustable: User could type name of program, so when that program starts cpu goes to full frequency or switches to faster governor settings, could be helpful (as it was with benchmarks) with most demanding games, making them run little faster. After the game finished phone would go back to it's normal power saving mode, win win for everybody.
At least they showed how inefficient on demand governor is and I for one will be looking into using different, better governor, to speed things up, since it seems for now the hardware is capable of better performance, but is artificially slowed down by poor governor programing.
 
Journyman16
Old
(Last edited by Journyman16; 8th March 2014 at 03:06 AM.)
#6  
Journyman16's Avatar
Senior Member - OP
Thanks Meter 208
Posts: 486
Join Date: Aug 2012
Location: Bendigo

 
DONATE TO ME
I used Governor tweak 'packs' on my S3 - there were performance tweaks, battery tweaks and more battery tweaks - after a ROM flash I'd run the 2 battery ones then the speed ones - my theory was the battery ones would reset everything to minimum then the speed one would reset only the ones needed for performance. Seemed to work, side by side with an S4 my phone was as fast (and faster in some things) and I got 2 to 3 days battery. (versus the S4 lucky to make a day )

Tweaks are the next thing to look at I guess...
Never doubt there is Truth, just doubt that you have it!
 
Reefee
Old
#7  
Reefee's Avatar
Member
Thanks Meter 30
Posts: 47
Join Date: Nov 2008
Location: Nottingham
I just tested it with 4.4 (X-Note 12 Rom by Solo and Xentrix , and Snapkat 2.2 kernel by Civato) and got 41K+ @2.88GHz CPU and 533MHz GPU both on performance governor. I don't have any 4.3 reference material unfortunately.
But I must say 4.4 feels much smoother than 4.3 and also better battery life.
Attached Thumbnails
Click image for larger version

Name:	2014-03-12 11.34.42.jpg
Views:	83
Size:	144.3 KB
ID:	2626730  
 
pete4k
Old
#8  
Senior Member
Thanks Meter 55
Posts: 326
Join Date: Jan 2014
I read somewhere KK uses interactive governor, is that true? Could that be one of the reasons phone seems more responsive?
 
djbigbear
Old
#9  
Junior Member
Thanks Meter 1
Posts: 23
Join Date: Dec 2013
exynos n900 user here.

using TR-Rom 4.4.2 kitkat
battery life is significantly longer.
antutu performance is boosted up from 33k to:

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes


TRENDING IN THEMER...