Benchmarksssssssss!!! - Galaxy S I9000 Q&A, Help & Troubleshooting

Hi guys!
can you post your SGS configuration and tell us how do you score with Quadrant Benchmark?
mine is that:
SGS i9000m
kernel speedmod 13c - rom ZeroGingerMIUI with no OC Lagfix ---> 1800
kernel speedmod 13c - rom ZeroGingerMIUI with 1.2ghz OC Lagfix---> 2100
kernel super optimized - rom ZeroGingerMIUI with 1.2ghz OC Lagfix--->2350
kernel hacked Supercurio for GingerBread - rom Ultibread v1.4 Lagfix(no system) --->2000
yes, i'm a bench_maniac!

Stock I9000 XXJVK --> ~awesome scoreon my user experience benchmark.

Awesome... no words to describe!

Quadrant doesn't reflect real life performance at all. I used to get 2200+ with a lagfix on 2.1 but it was no where close to the real life performance of what I experience now on 2.3.3 JVK with no lagfix where Quadrant score is just 1020 .

Nice
Sent from my GT-I9000 using Tapatalk

quadrant doesn't reflect the actual user performance/experience because in the past i've had high quadrant scores but poor user experience.
but as the versions change (i.e eclair>froyo>gingerbread) i get lower scores but better user performance.

Agree my user experience benchmark is also off the scale .
Quadrant crap as always .
Why dont you post all these benchmark quadrant posts in the same place if you want to compare size.
jje

1200 on miui
Sent from my GT-I9000 using XDA App

Mine is scoring "awesome" too in user experience!

I personally think this thread should be removed I'm afraid.. For starters, there are a billion other quadrant threads, and nobody has F***king bothered, as mentioned to collate the values into a table. Also, nothing is worse than breeding a culture of people who believe that Quadrant accurately measures the mobile performance for all workloads.

Auzy said:
I personally think this thread should be removed I'm afraid.. For starters, there are a billion other quadrant threads, and nobody has F***king bothered, as mentioned to collate the values into a table. Also, nothing is worse than breeding a culture of people who believe that Quadrant accurately measures the mobile performance for all workloads.
Click to expand...
Click to collapse
I agree with Ausy that this kind of thread is not of much practical use. I run Quadrant weekly, but only to compare my phone to itself as a diagnostic. If the Q score changes radically, I try to find out why. I also use it to compare ROMs.
However, I don't see that the benchmark scores of two different phones will tell you anything of much value when you compare them.

Related

Guys/girls with froyo

Coming over from fascinate forums and just wondering how the benchmarks test change when we actually get some froyo. I know they don't matter but it helps show the overall boost in performance so anybody with linpack or quadrant if u don't care please post ur results. Thanks n advance. Don't flame me, this will give u cappy users a chance to brag
Sent from my SCH-I500 using XDA App
Quadrant scores stay pretty close to the same, mid 900's to low 1000's, linpack clocks in between 13.5 and 15
As has been mentioned in many other threads, Quadrant scores are biased toward Snapdragon processors.
These benchmarks really don't mean anything from one phone to another, but within the same phone model it does give an idea of improvement..
Sent from my SAMSUNG-SGH-I897 using XDA App
1842 Quadrant Score after OCLF on Froyo which is great compared to 900 before the fix
This is with RyanZA’s EXT2 Lagfix.
Full Quadrant: 1867 (Saw 67 FPS at one point) This is substantially lower than the 2.1 EXT4 Lagfix, but still twice as high as the stock score of 957. I know Quadrant really doesn’t mean anything.
Benchmark PI: 1445 MS
Linpak:
MFLOPS: 13.809
Time: 6.07 Seconds
Norm Res: 5.68
Precision: 2.22044
The "Billabond1 real life use" benchmark shows performance to be as good or better than 2.1. I went from cognition 2.1 w/ voodoo lagfix to cognition 2.2 w/ oclf. I would say that the performance is about the same, good, and the features in 2.2 make it totally worth the update.
On a fresh install of 2.2 I got 995 on my quadrant score.
When I finished rooting and putting on RyanZA's OCLF, my quadrant went up to 2105.
Never did a linpack

[Q] Quadrant CPU Score Nexus S vs SGS

So a few news sites are showing off the Quadrant score of the Nexus S running CM7: http://www.talkandroid.com/27099-nexus-s-scores-high-on-quadrant-benchmarks-running-cyanogenmod-7/
One thing I noticed is that the CPU Score is 5372. When I run Quadrant Advanced on my i9000 the CPU Score is only 1371. What has changed in CM7 that gives the Hummingbird processor so much more computational power and how do we get it on the SGS?
Think those guys just know what they are doing when it comes to Android unlike Samsung :/ lol
I'd be interested to see how the stock Nexus S firmware does in the same benchmark. I don't care about the overall Quadrant score, just the cpu part. I wish we could have CM on SGS

[Q] Tips to increase Quadrant score.

Cureently i'm using Darkky's ROM 8.0 with Advanced Voodoo kernel v666.2. The present score is 1500 with lag fix. How to improve the scores further ??
why do you even want to get high score??? you can install another kernel to overlock processor
Sent from my GT-I9000 using XDA App
Best method is to lie .
Quadrant scores are irrelevant and bear no relation to the speed of the phone since Eclair .
Reason lagfixed Eclair posted high scores . Far faster less laggy Froyo posts lower scores = Quadrant no test of Froyo .
jje

[Q] Linpack for android..am i lacking?

Ok so i flash theses beautiful roms and now ive tried some testing and currently on my captivate i ran the "linpack for andoid" app and it states:
MFLOPS: 13.982
time: 6.0
Norm Res:5.68
Precision: Inaccurate result
ok so i looked at the top 10 section ounder my phone and their MFLOPS is between 20-50 MFLOPS...So do i need some tweaking?..I mean the phone is good , Rom is awesome...any thoughts?
Captivate 1007
Phoenix 5.2 (2.2.1 JS5)
Speedmod i897 500hz k13c
UGJK4
Lagfix enabled
(r) Tweaks enabled
I wouldn't worry about it. First, your test said inaccurate results, so try running it a few times, though you probably won't get much more out of it. Secondly, most people believe these benchmarks are a load of BS, and I would trends to agree. If the ROM feels snappier and works smoother than stock for you, there really isn't any use in the benchmark scores. I'm running Serendipity 5.11 with the same kernel as you, and I find my scores are in the low 14's- which, as is for you, seems slow compared to the online results. However, my phone supposedly blows the other results out of the water in the Quadrant Standard benchmark, regularly achieving scores over 1500, and I imagine yours would perform similarly, so no, I wouldn't worry at all.
Overall, most of the roms on here achieve faster performance along with better battery life compared to a stock cappy. I suggest trying out as many roms as possible to find the one that fits you best, as everybody has different tastes, and each rom behaves differently for everybody.
Personally, I think Serendipity is something you should try, as it looks friggen sweet as is, and even better IMO with some tweaks. I changed the battery indicator to the android dude, the font to Armani, and use launcher pro over the base rom. Altogether, my phone is super slick and snappy. Oh yeah, it also accesses HSPA networks, so downloads are way quicker than stock (pretty sure your current rom does too).
Sent from my SAMSUNG-SGH-I897 using XDA App
that's about right, repeated tests should get you 14+. it does say inacurate so i dont think you have a problem, just a process running somewhere. the hummingbird lacks a bit in this area and snapdragons excel only in this area. i know it is weird that the nexus one and droid incredible smoke this phone in linpack but it is hardly a thorough benchmark. run quadrant on CM7 when it comes out and see what the hardware is really doing without the samsung influence. despite the low mflops you will see good scores across the board.
Thanks alot for the input everyone...jus curious how gen-pop was adding up... but who cares about the numbers i suppose when my phone is soo [email protected]## cool hahha
Sent from my GT-I9000 using XDA App
Just for the record, I'm on stock Froyo, With no SpeedMod, but with voodoo lagfix and I got better benchmark, linpack and quadrant scores than I did with a custom rom and Speedmod kernel before. Add to that the stability of a native ROM and kernel, and I think you know what I am saying here...
clemmie said:
Just for the record, I'm on stock Froyo, With no SpeedMod, but with voodoo lagfix and I got better benchmark, linpack and quadrant scores than I did with a custom rom and Speedmod kernel before. Add to that the stability of a native ROM and kernel, and I think you know what I am saying here...
Click to expand...
Click to collapse
Hmmm. Intresting. Makes sense. Whered ya get the stock from..curious about the release tomorrow for At&t. Think i might try some configs with it.
Sent from my GT-I9000 using XDA App

Benchmark-Scores low

Hi there!
I'm new to the Optimus 2x rom stuff. I used a Galaxy Spica before, which I rooted and installed some nice roms on.
On my LG it seems that the custom-roms I chose so far make no difference in speed compared to the stock rom.
Quadrantscore: ~2200 with WIUI 2.5.25 Stable (GB) and ETaNa STOCK_LITE 2.6.32 which is (according to Quadrant) below the O2X.
In some rom-threads I see scores of about 4000-6000 how do I reach that? Did I do something wrong?
greetings, and thanks for your tips
reijin
//PS.: I'm aware, that those scores are not a prefect representation of how the overall experience with a Rom is, but I'd like to know. At least the new Rom feels way better then the original stock rom.
Yeah it's a bit strange. I'm using cm7 on mine, the general performance is really nice and fast no complaints from me but running quadrant benchmark I get scores of around 2200-2300. My friends o2x hits scores closer to 3000 on the same rom. Both using temasek's kernel too.
Sent from my MB860 using xda app-developers app
I only have experience with ICS ROMs and kernels, but I can testify I've consistently had quadrant scores >4000. It's a complicated mix of the hardware you have (no two o2x's are the same), whether your kernel allows overclock, how optimized the io is, etc.
Sent from my LG-SU660 using xda premium
try to use antutu benchmark and see your score. and you're using stock lite which is 1.0ghz the clock. if you want to get higher score try HOC lite, and use OC/UV programs to control your clock cpu. i use pimp my cpu profile i set:
- daily freq: 216-1015MHZ.
- bench freq (like antutu,quardrant, necore, vellamo, and so on): 216-1418MHZ.
- and so on, what ever profile you want
You used spica and you don't know this basics?
Spica's angry anboys can tell you.
I can say that high quadrant scores can be resulted when IO rate is very high. You can cheat quadrant at it will think IO rate is high but it won't make phone a formula 1.
Use app2ad or something that uses ext partition and then run quadrant.
I am afraid y used spica and didn't know it.
Sent from LG 2X using Tapatalk 3 Beta
With CM9 Harsh kernel i have in Quadrant 3200 points

Categories

Resources