Database Users

hey flipz how about another benchmark

how bout another benchmark for all these latest and greatest roms that claim to be too fast for class. so we know the truth once again

I agree
Sent from my HERO200 using the XDA mobile application powered by Tapatalk

No problem. I should have time to do it tomorrow.

flipzmode said:
No problem. I should have time to do it tomorrow.
Click to expand...
Click to collapse
haha, awesome...

flipzmode said:
No problem. I should have time to do it tomorrow.
Click to expand...
Click to collapse
wow awesome quick response too... i knew i liked u for some reason

Won't hardware benchmarks always be the same between all ROMs?
Is there a way to objectively benchmark resource management? Speed of the interface, opening apps etc.

i didnt see the last one but id like to see benchmarks like on te gsm roms
http://forum.xda-developers.com/showthread.php?t=656818
Also the following for benchmarking:
For 3D Performance
* Qualcomm's "NeoCore" 1.9.18 - Measured in FPS
For 2D Performance
* Softweg's "Benchmark" 1.03 - Measured by an internal score
* "Lag test" - Originally to measure 2D performance for Motorola Droid users. Measured in FPS
For CPU Performance
* Linpack for Android - Measured in MFLOPS
All of the tests mean: The higher, the better.
The method of testing was the usual flashing, a couple of reboots to make sure every caching necessary was done and the average of 3 benchmark of each test.
Click to expand...
Click to collapse

Make sure to test Pancake 0.1! Still the fastest rom out there .

bikcmp said:
Still the fastest rom out there .
Click to expand...
Click to collapse
I really liked your rom, but seriously....
;/

TrevE said:
i didnt see the last one but id like to see benchmarks like on te gsm roms
http://forum.xda-developers.com/showthread.php?t=656818
Click to expand...
Click to collapse
http://geekfor.me/news/benchmarking-sprint21/
I'll add linpack to the list too.
bikcmp said:
Make sure to test Pancake 0.1! Still the fastest rom out there .
Click to expand...
Click to collapse
And it's also 1.5 which makes it irrelevant for these tests.

Someone should come out with a rom, with JIT enabled.
Then they can claim they have the fastest rom.
I had it running on dc 2.7.2 without any problems.. Just cant use anything HTC, like lockscreen or sense.
But nexus or helix launcher and beautiful widgets, everything is normal again.

JIT is too buggy still, im looking foward to playing with compcache when someone gets 2.6.29 source working
http://wiki.cyanogenmod.com/index.php/Compcache
thanks flipz just looked it over like the benchmark setup

TrevE said:
JIT is too buggy still, im looking foward to playing with compcache when someone gets 2.6.29 source working
http://wiki.cyanogenmod.com/index.php/Compcache
Click to expand...
Click to collapse
Yea. We had some goodies starting to go in the .27 kernel but then the 2.1 stuff became the main focus so it never got finished. Ha.

TrevE said:
JIT is too buggy still, im looking foward to playing with compcache when someone gets 2.6.29 source working
http://wiki.cyanogenmod.com/index.php/Compcache
Click to expand...
Click to collapse
I agree its not perfect.. But i can use it on my daily rom and run every app I would on a normal day without an issue.
Before dc 2.06 came out, i ran it on dc 2.05 for 3 days straight.'
I dont use bluetooth or some other random things. So it would probably start 101 problems for some people.. Like theres not enough already.

Whats the fastest custom rom?

yeah whats everyones fastest most stable combination?mine so far would be nfinitefx45 deoxed stock rom with a 768 overclocked kernal.and the new CyanogenMod-6 rc is pretty fast!with the uncapped kernal

Cyanogen 6 with dec s uncapped kernel clocked to 768. Using battery tweak, juice defender, jit enabled, surface dithering disabled, home locked in memory. I get about 420 on quadrant and around 5.2 mflops
Sent from my HERO200 using XDA App

cyanogen
i recently installed cyanogen 6 with the google addon.
What is the " with dec s uncapped kernel clocked to 768. Using battery tweak, juice defender, jit enabled, surface dithering disabled, home locked in memory" you are refering to?
I currently use set cpu to overclock to 691. I currently am only getting just under 3 mflops and just under 400 on quadrant.

jblair496 said:
i recently installed cyanogen 6 with the google addon.
What is the " with dec s uncapped kernel clocked to 768. Using battery tweak, juice defender, jit enabled, surface dithering disabled, home locked in memory" you are refering to?
I currently use set cpu to overclock to 691. I currently am only getting just under 3 mflops and just under 400 on quadrant.
Click to expand...
Click to collapse
Decad3nce has an uncapped Kernel - the stock cm6 one is capped so users with older chips (without the "With Google" branding) can boot.
Battery Tweak by Dec and Collin_Ph improves battery life by installing a small user defined service (I can't remember what exactly it does)
Juice defender is an app in the market that controls battery use of apps.
JIT - "Just In Time" compiling - very techy stuff here but more or less gives you more MFLOPS - but our Heros dont have enough ram for it to truely be effective.
Surface Dithering is a CM6 option that makes the screen look better. Google the term and see what it does for photos and stuff.
Home Locked in memory keeps the OS from flushing it from cache and causing a few issues, like hitting the home button and all the screens being blank for a second while it populates them.
So the above statement is just his lits of configs that make the device run best for him. But remember MFLOPs and Quadrants aren't everything Alot of it is feel.

Wow that was nice and detailed explanation. I'd up vote it if I could.
I haven't used the uncalled kernal. I have with google branding. I think I'll try it anyway.
Cm6 fresh install seems faster than any other rom I've ever tried. I turn off eye candy and it flies
Sent from my HERO200 using XDA App

lol
pancake 0.1 jk
RegawMOD 2.1t

regaw_leinad said:
lol
pancake 0.1 jk
RegawMOD 2.1t
Click to expand...
Click to collapse
Damn Regaw is on the ball lately.
Haven't seen you post in a while. Still in school?

My phone seems to be handling uncapped kernal quite well after all
Sent from my HERO200 using XDA App

Kcarpenter said:
Damn Regaw is on the ball lately.
Haven't seen you post in a while. Still in school?
Click to expand...
Click to collapse
haha, yeah I just got back into this again from that damn break. It's still summer time, but I'm playing on a cruise ship for a year starting in a few weeks.

regaw_leinad said:
haha, yeah I just got back into this again from that damn break. It's still summer time, but I'm playing on a cruise ship for a year starting in a few weeks.
Click to expand...
Click to collapse
I've though of doing that from time to time but don't seem to be suited for any of the jobs they offer Americans at decent wage. I'm best suited for engine room with my navy training. Reason to go with cruise line would be to see nicer destination ports
Sent from my HERO200 using XDA App

I gave up using Fresh ROM because it was getting laggy and we haven't had an update since August, so I finally made the switch to Cyanogen's RC1. Overclocked to 768 mhz, I've been getting 438 consistently on Quadrant. Linpack has been giving me varying results around 3 MFLOPS. The best I've gotten was 3.25.
I'm really curious as to how other Heroes do!

jcha said:
I gave up using Fresh ROM because it was getting laggy and we haven't had an update since August, so I finally made the switch to Cyanogen's RC1. Overclocked to 768 mhz, I've been getting 438 consistently on Quadrant. Linpack has been giving me varying results around 3 MFLOPS. The best I've gotten was 3.25.
I'm really curious as to how other Heroes do!
Click to expand...
Click to collapse
Switch to CM 6.1.0, very fast and stable

where can we find this Home Locked in memory app?
Kcarpenter said:
Decad3nce has an uncapped Kernel - the stock cm6 one is capped so users with older chips (without the "With Google" branding) can boot.
Battery Tweak by Dec and Collin_Ph improves battery life by installing a small user defined service (I can't remember what exactly it does)
Juice defender is an app in the market that controls battery use of apps.
JIT - "Just In Time" compiling - very techy stuff here but more or less gives you more MFLOPS - but our Heros dont have enough ram for it to truely be effective.
Surface Dithering is a CM6 option that makes the screen look better. Google the term and see what it does for photos and stuff.
Home Locked in memory keeps the OS from flushing it from cache and causing a few issues, like hitting the home button and all the screens being blank for a second while it populates them.
So the above statement is just his lits of configs that make the device run best for him. But remember MFLOPs and Quadrants aren't everything Alot of it is feel.
Click to expand...
Click to collapse

Jit, surface dithering and home locked are settings in cm6. Not apps. I have cm6 vanilla and tried extremely blue Rom, best things ever. And constantly updated. It's nice to know they care. Thanks to all u devs by the way

Benchmark Tests on Legend 2.1 vs 2.2

Hi All,
Updated my missus's Legend to 2.2 today and ran a Quadrant Benchmark test on it..
best of 3 was 903 (avg is about 860)..
What are you getting on 2.1... and what sort of scores are you getting on the custom ROM's out there?
Post your scores stating which ROM you're running..

Clean Eclair, 2.1 = 4XX-5XX
Azure, 2.2 = 8XX-9XX
Cyanogenmod RC2, 2.2.1 = 8XX-9XX
Never benchmark a Sense ROM before so I cannot provide such information.

Best of 3 was 365 on unbranded 2.1 update 1
Sent from my HTC Legend using XDA App

Constantly above 1xxx with CM RC1/RC2 and OC787 kernel...but that's normal since it's overclocked
Sent from my Legend using XDA App

My result was 387 on 2.1 update 1 official Rom
Sent from my HTC Legend using XDA App

im_iceman said:
Hi All,
Updated my missus's Legend to 2.2 today and ran a Quadrant Benchmark test on it..
best of 3 was 903 (avg is about 860)..
What are you getting on 2.1... and what sort of scores are you getting on the custom ROM's out there?
Post your scores stating which ROM you're running..
Click to expand...
Click to collapse
Using a normal 2.1 legend, benchmarked three times... 377, 332 and 361. Even more anxious to get froyo now then!

rajasyaitan said:
Azure, 2.2 = 4XX-5XX
Cyanogenmod RC2, 2.2.1 = 8XX-11XX
Click to expand...
Click to collapse
Fixed....
Sense 2.1 was 3XX (4XX with OC)
However, even going from 3XX to 8XX (CM6.1) I did not see any speed improvements in games. OC DID improve the speed

Skardy said:
Fixed....
Click to expand...
Click to collapse
You cannot get 11XX without OC'ing.

rajasyaitan said:
You cannot get 11XX without OC'ing.
Click to expand...
Click to collapse
Right I forgot to specify

First run on rooted Sense Froyo was 915. Second run was 870. Third was 927.

rajasyaitan said:
Clean Eclair, 2.1 = 4XX-5XX
Azure, 2.2 = 8XX-9XX
Cyanogenmod RC2, 2.2.1 = 8XX-9XX
Never benchmark a Sense ROM before so I cannot provide such information.
Click to expand...
Click to collapse
I'm sorry Sir, but are you ****ing-kidding-me? With almost stock 2.1 i had ~380. With azure 1.0.x now i have ~460.

I have not programmed anything in android yet, but I know programming java for some years and I think that I can tell you why those spectacular benchmarks doesn't apply in most games.
The improvements have been made in the dalvik virtual machine, which is responsible for executing the java code of the applications.
But if programmes want more performance, they can write part of the program directly in native code, developed in C or C++. This code cannot be optimized (more), thus applicacions using JNI massively don't see any improvement.
I suppose many games should be using JNI because of the advantages in performance, so they're already 'optimized'.
Please, correct me if I'm wrong
--Edu

no0neX said:
I'm sorry Sir, but are you ****ing-kidding-me? With almost stock 2.1 i had ~380. With azure 1.0.x now i have ~460.
Click to expand...
Click to collapse
I'm sorry Sir, but are you a ****ing kid? Let me try to be as nice as possible here. Do your homework before you bash other people. Go to the respected ROM's thread, or the OC thread and look around for other peoples results. That is the average result. 4XX on Azure? Seriously? Throw your phone to the wall and maybe that'll boost the Quadrant scores.

I'm with you my friend. Some people are just ignorant.
rajasyaitan said:
I'm sorry Sir, but are you a ****ing kid? Let me try to be as nice as possible here. Do your homework before you bash other people. Go to the respected ROM's thread, or the OC thread and look around for other peoples results. That is the average result. 4XX on Azure? Seriously? Throw your phone to the wall and maybe that'll boost the Quadrant scores.
Click to expand...
Click to collapse
Sent from my Legend using XDA App

rajasyaitan said:
I'm sorry Sir, but are you a ****ing kid? Let me try to be as nice as possible here. Do your homework before you bash other people. Go to the respected ROM's thread, or the OC thread and look around for other peoples results. That is the average result. 4XX on Azure? Seriously? Throw your phone to the wall and maybe that'll boost the Quadrant scores.
Click to expand...
Click to collapse
Raja, the avarage Azure scores are 4XX because JIT is actually disabled in that rom, even though the option present and checked.

Skardy said:
Raja, the avarage Azure scores are 4XX because JIT is actually disabled in that rom, even though the option present and checked.
Click to expand...
Click to collapse
Actually it is enabled. I asked Ali about it before. But then again that was few months ago, I might be wrong.

rajasyaitan said:
Actually it is enabled. I asked Ali about it before. But then again that was few months ago, I might be wrong.
Click to expand...
Click to collapse
Sorry my bad.
Seems like JIT was disabled only until Azure 0.6.
However I can confirm that my Azure 1.0.1 quadrant score was 4XX (which is kind of low...), and as I now have 1095 with CM6.1 I don't think theres a problem with my legend's performance.

Sorry for the newbie question, but.. how do I run a Quadrant Benchmark ?
I know these numbers refer to the performance, and that the 2.2 scores a lot better, but in actual everyday usage - battery life, speed etc. what does this practically mean?

Skardy said:
Sorry my bad.
Seems like JIT was disabled only until Azure 0.6.
However I can confirm that my Azure 1.0.1 quadrant score was 4XX (which is kind of low...), and as I now have 1095 with CM6.1 I don't think theres a problem with my legend's performance.
Click to expand...
Click to collapse
No worries mate. Seriously? The last time I used Azure I got around 8XX. That was version 1.0 I think.
versifier said:
Sorry for the newbie question, but.. how do I run a Quadrant Benchmark ?
I know these numbers refer to the performance, and that the 2.2 scores a lot better, but in actual everyday usage - battery life, speed etc. what does this practically mean?
Click to expand...
Click to collapse
Install Quadrant and run it. Well, technically it is just numbers. But most of the benchmarkers do admit feeling the difference between lower and higher scores. Some say those are placebo effect, but meh. I don't buy that ****.

2.1 Stock
1st test
380
2nd test
381
3rd test
383
/T

[Q] No JIT in DK28

So DK28 is supposedly 2.2.1 but I ran Quandrat with the stock rom and it comes up 873, almost exactly the same as Galaxy S 2.1, leading me to believe that JIT is not in the rom. Correct?

avenger213 said:
So DK28 is supposedly 2.2.1 but I ran Quandrat with the stock rom and it comes up 873, almost exactly the same as Galaxy S 2.1, leading me to believe that JIT is not in the rom. Correct?
Click to expand...
Click to collapse
I may be wrong but quadrant:
1) Isn't the best way to base performance
2) Doesn't really show improvements that JIT provides
Download linpack if you want and your score should be around a 14 or so, compared to the 8.5-9 on 2.1

Quadrant isn't a real life benchmark, and is highly flawed. Check out Quantum's "warp speed edition" ROM for proof.
Sent from my SPH-D700 using XDA App

I got 1205 with stock dk28 but I have noticed others reporting lower scores. Variability in testing with a bad benchmark perhaps?

I've got 1839 lol... For some reason I doubt that's accurate

I did a quadrant test with a stock epic 2.1 and an epic on dk28 and the scores were pretty much the same every time. The major difference was that the Froyo epic ran the cpu benchmarks much faster but hung during the first I/O test. The 2.1 epic ran everything about the same speed and completed the whole test much faster than the froyo epic.

I also noticed this and posted in another thread...
to quote myself
I was getting around 8.3xx in 2.1, and now I'm getting 13.7xx in 2.2.
(I used the 1506bdf2e04b.update-SPH-D700_DI18ToDK28.zip renamed to update.zip method.)
Saw a video the other day showing a nexus 1 scoring 20+ on 1st run and 40+ on 2nd.. My friends Evo (with 2.2) scored @ 38 or so..
Quadrant still shows me well under all of the 2.2+ phones listed.
Click to expand...
Click to collapse
I expected a bit more of an increase..
I too am wondering if the JIT was left out of our Froyo (if that's even possible, I'm still a total nube to andriod stuff).

What exactly is JIT, does it speed the phone up?
Sent from my SPH-D700 using XDA App

Just In Time compiling...
From what I know about .Net (I'm a C# coder), it allows for compilation of the app for the specific CPU during the 1st run.
Someone please correct me if this differes from andriod's JIT...

the epic has fast computation hw between the cpu and gpu, it could be the case that the JIT benefit doesn't show in the benchmark margins.
the device is a fast one, all compared.

MFLOPS

2.1 got 7 mflops,2.2 roms got 5 mflops and 2.3 roms 8 mflops but looks slow because 2d is not compiled from source.
Then what rom is fastest?

Well, 2.1 gets 7MFLOPS in linpack, but no more than 400-500 points in quadrant advanced, while 2.2 gets those 5-5.5 points in linpack but 1200 and above in quadrant advanced, if OCed. Now, what's the conclusion? Don't look at benchmarks, look at user experience.
Truth is, eclair and froyo are performing quite similarly. Some games lag in both, some in only on distribution. As far as homescreen and menus go, I feel that my GO launcher EX runs more smoothly on froyo than on eclair, while froyo is even more loaded with apps then my phone was back in its 2.1 days. And bear in mind that it's a "heavy" launcher with lots of additional features, like app killer.
So, the conclusion is, that froyo may in fact be a bit faster, but not much, and it may be due to better memory management of a newer android version. What is important, we get native app2sd, CM settings and other cool stuff if we use a custom ROM.
As far as 2.3 distributions are concerned, I cannot comment, as I never had one for more than one day, they were either laggy, or had no camera, or had other bugs. I'll wait 'till a functional one comes out, then I'll test it extensively