Database Users

Spikeys ROM benchmarks (3/3/10 - added Cyanogen 4.2.15)

I know some people don't believe in benchmarks for Android, so take it for what it's worth. To me, the results have been somewhat on point, in terms of the speed of the ROM. I will be updating this as more ROMs come out, but if anyone wants another ROM tested, just let me know.
About the benchmarks...
- All test were done after a clean flash of the ROM.
- ROM was never signed into Google.
- Only apps installed were Advanced Task Manager, SetCPU and Swapper(and benchmark apps).
- All test were done while in Airplane Mode(to cut down on background network activity).
- Ended all processes before running each test.
- Alls ROMs(except the Hero ROMs) had Swap/Compcache disabled and CPU set at 384min/528max.
- Hero ROMs were set with 96MB swap and CPU set at 384min/528max.
- The results here are from the best of 5 runs. I did not want to take the average because there were some inconsistent results that would have thrown it off.
Lower = Faster only for the first test.
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}

nice. Can't say I would of had the patience to do this, lol. Very interesting to see though, thanks for sharing.

Klyentel said:
nice. Can't say I would of had the patience to do this, lol. Very interesting to see though, thanks for sharing.
Click to expand...
Click to collapse
It really doesnt take that long. Just flash the new test ROM, batch install a few apps, run the test, then nandroid back to my everyday ROM.

Updated
10char

Is faster lower on all 3 tests or just on the first one?

billygoat555 said:
Is faster lower on all 3 tests or just on the first one?
Click to expand...
Click to collapse
Just the first one is lower=faster.

Nero 2 with Super I/O Kernel, Yes or No?

will it make difference?

I just finished my titanium restore, but it's smooth so far.
A lot of nice little touches. And Quadrant says 1193 so that ain't bad.

{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
Runnin nero+dow+oclf

im only running around 1000 with nero 2 and IO kernel... not sure why

willsnews said:
Runnin nero+dow+oclf
Click to expand...
Click to collapse
What is dow?

Good god, yes, this is a terrific combination. Buttery smooth.
As for varying Quadrant scores, I find that if I kill all user apps right before running the benchmark, I'll hit in the 1200 range. Otherwise I'm only around 1000. Doesn't matter either way though, because it's silky smooth either way. I haven't even bothered with a lagfix because I don't think it's necessary. I noticed that TW tweaked the mounts so that even though it's still RFS, it's considerably faster than stock JK2 or JK6. Eugene's kernel is also the bomb and I've seen great battery and stability.
I've been waiting for a while for a Froyo build that approximates my reference setup on Eclair, which was the final Fusion w/ OC/UV (1.2ghz) with z4mod. That is the benchmark for me, and I think I've finally found it's equal in Froyo. Very happy with my Vibrant right now.

I completely agree this is the first froyo rom that matches that speed and stability I had with fusion oc'ed lag fixed.
Sent from my SGH-T959 using XDA App

ionic7 said:
What is dow?
Click to expand...
Click to collapse
Destroyer of worlds
Runnin nero+dow+oclf

quadrant score on revolution hd 2.0.11

i think i have a problem.
i can only get a max of 1800 points in quadrant standard with revolution hd rom and setcpu with autodemand 245-1228.
why is this happening? i've done a factory default and the same.

Don't pay attention to these benchmarks... they tell you nothing. And noting wrong with that score, it's high enough, upgrading to ARHD 3.x might help.
If you are dumb enough to rate a phones performance on such bad benchmarks then..
Turn setCPU to performance
Restart your phone and kill all apps with task killer
Turn off sync/wifi

and update your rom maybe..
AR 2.0.11? AR 3.3 is already out...

quadrant scored really doesn't say a lot for the ROMS. you may have high 2000s scores but your battery lasts only have a day.. doesn't really means a thing

For me it was difficult to get consistent results with Quadrant.
Upgrading the app changed results dramatically.
You can run a benchmark now, and again in 5 minutes time and the results can be way different.
It was great at first but frustrating after a while.
For example: (on stock rom)
Original (2053)
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
Then I updated the app (1773)
Then 12 minutes later (1813)
Obviously many factors are involved but as Foladko and Baste07 suggest - I wouldn't pay too much attention. It's not as cut and dry as for example overclocking a desktop PC.

[quadrant scores]

I know is basically meaningless but still post your top score
just for grins
heres mine\/\/\/\/\/
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
CM7
Netarchy-nexus 1.3.0.2
max cpu 1300
min cpu 800
OC'd

miui 1.4.15
oc 1,4ghz performance

wuadrant scores actually do have meaning as long as you are comparind scores within the same device hardware (and not using any cheats or tricks) because it lets you compare how well your hardware is performing based on other people's harddware. but it does NOT have meaning comparind scores across different device models. at that point its meaningless.
company A could optimize a 500mHz phone to be smooth as butter, super optimized and efficient working on it for 3 years. while company B could not do crap, release a 2gHz phone, but not optimize it at all. so scores between those 2 devices would be meaningless.

RogerPodacter said:
wuadrant scores actually do have meaning as long as you are comparind scores within the same device hardware (and not using any cheats or tricks) because it lets you compare how well your hardware is performing based on other people's harddware. but it does NOT have meaning comparind scores across different device models. at that point its meaningless.
company A could optimize a 500mHz phone to be smooth as butter, super optimized and efficient working on it for 3 years. while company B could not do crap, release a 2gHz phone, but not optimize it at all. so scores between those 2 devices would be meaningless.
Click to expand...
Click to collapse
theres no way to distinguish if they're cheating on the benchmark.. unless they're using Quadrant Advanced.

[OCT20] ICS Sense Kernel OC comparison.

There are two current ICS Sense Kernel for DHD, as what I know.
BeastMode
Virtuous
I had comparison with few benchmark tools.
wifi on, mobile off, auto-sync, gps ..etc are off.
max freq selected most near to 1.5GHz (1.498 or 1.517GHz). governor performance(always Max freq). none altered.
omittted first result. these are 2nd and 3rd. also scores could vary with configurations, apps, receptions etc. this is just n00b's comparison.
While benchmarking, No FC or error found except Linpack for Android.
(found Inaccurate Result in both 3 of them not always.. wish It had no advertisement while "Testing.." ad could poison this.)
Stability and Battery Life is not in my concern.
needed to save battery from wear is highly expected.. its 2yrs old now..
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
site resized that chars may visible. numbers with red are Average.
BeastMode1.5 linaro compiled by p2pjack
BeastMode1.5 gcc compiled by p2pjack
Virtuous 3.0.43 Kernel by chrisch1974

Beast Mode Kernel links are dead mate. any chance of re linking please.

Because those are posted on Sabsa Prime thread, and it was invisible for a while. Now links are back!.
Sent from my HTC Desire HD using xda app-developers app