Database Users

Benchmark PI

Hello guys, just wanted to share my results with you: (Average after 5 runs)
JM2: 2576
JP2: 1455
JP3: 1503

Arkymedes said:
Hello guys, just wanted to share my results with you:
JM2: 2576
JP2: 1455
Click to expand...
Click to collapse
Could you share some more info with us?
How did you achieve these numbers?

Impressing. I guess we have to live with the fact that JP2 is faster. Hope it isn't too buggy then..

ivas75 said:
Could you share some more info with us?
How did you achieve these numbers?
Click to expand...
Click to collapse
I think there is nothing much to share I guess.
I just flashed the ROM, start Benchmark PI from Market and posted the scores.
The Froyo build, even alpha the way it is already makes the difference. And quite a difference.

Edited first post

JP2 samset 2.0 - 1302
Sent from my GT-I9000 using XDA App

have you tried running it with "Linpack mark"
that is what most people use to test the smartphones speed, rather than the one you are using
i think both are good

samsung is improving RFS! proof inside!

I used quadrant with ramdisk hack and without.. the answer shocked me:
froyo dk05 with ramdisk hack: 2200~
eclair di18 with ramdisk hack: 2400~
froyo with latest 2.6.32.9 dk05 kernel with rfs: 1100~
eclair with 2.6.29 di18 kernel with rfs: 900~

Froyo
I have a diffrent froyo than you mine is froyo dj29 how and why is yours diffrent

josephrr21 said:
I have a diffrent froyo than you mine is froyo dj29 how and why is yours diffrent
Click to expand...
Click to collapse
noob is one of the group of Devs that are getting the leaks before they are released to us.
He is referring to a later, and hopefully less buggy, froyo build. According to screenshots posted elsewhere, that build build also appears to have the modem.bin for the phone.
I am sure that they will release a newer build when they are able and after they are sure it is reasonably stable.
We do not need a forum full of bricked Epics, do we?

What is the ramdisk hack and does it provide any real benefits?

noobnl said:
I used quadrant with ramdisk hack and without.. the answer shocked me:
froyo dk05 with ramdisk hack: 2200~
eclair di18 with ramdisk hack: 2400~
froyo with latest 2.6.32.9 dk05 kernel with rfs: 1100~
eclair with 2.6.29 di18 kernel with rfs: 900~
Click to expand...
Click to collapse
pardon my ignorance but what is that suppose to mean?

yeah same here, what does that all mean? lol....
and 2200,2400 quadrant scores are the highest i've ever heard of

Eazail70x7 said:
What is the ramdisk hack and does it provide any real benefits?
Click to expand...
Click to collapse
cocochannel said:
pardon my ignorance but what is that suppose to mean?
Click to expand...
Click to collapse
razorseal said:
yeah same here, what does that all mean? lol....
and 2200,2400 quadrant scores are the highest i've ever heard of
Click to expand...
Click to collapse
If it's anything like a ramdisk on a computer,
it's setting aside RAM to use as storage (program install) space.
That is why the benchmarks are much higher than normal... the speed of file access goes way up.
It isn't worth doing to 'improve' your scores as it isn't doing anything practical. If you wanted to make a ramdisk for a couple of VERY frequently used programs, that might be a different story... but that might even require reinstallation after every reboot. Once a program 'loads' into memory it is usually getting 90% of the benefit installation on a ramdisk would provide.

what it means everything is slower in froyo except for i/o performance

noobnl said:
what it means everything is slower in froyo except for i/o performance
Click to expand...
Click to collapse
So is this a good thing or bad thing? Not sure...

Eazail70x7 said:
So is this a good thing or bad thing? Not sure...
Click to expand...
Click to collapse
a good thing, they are fixing rfs

noobnl said:
what it means everything is slower in froyo except for i/o performance
Click to expand...
Click to collapse
ouch
would you rather have either or?
as in what would our epic benefit more from

hayabusa1300cc said:
ouch
would you rather have either or?
as in what would our epic benefit more from
Click to expand...
Click to collapse
He answered it on there previous page.

noobnl said:
a good thing, they are fixing rfs
Click to expand...
Click to collapse
RFS is the Samsung "Robust" FAT File System? What is the metric (units)? A +20% bump looks good on a forum thread.
So this means that the system under load in froyo is ~8% off peak capacity vs elcair but the default system I/O performance is up ~20%
Please forgive the newb query - not as familiar with the inner workings.
<--- still patiently waiting for an official stable release

Great Info! Thanks

codest3r said:
RFS is the Samsung "Robust" FAT File System? What is the metric (units)? A +20% bump looks good on a forum thread.
So this means that the system under load in froyo is ~8% off peak capacity vs elcair but the default system I/O performance is up ~20%
Please forgive the newb query - not as familiar with the inner workings.
<--- still patiently waiting for an official stable release
Click to expand...
Click to collapse
yup .......

Woot. Bothers me to see people with slower phones trying to shove benchmark scores in our face.
Part of it is Quadrant's fault, really.

I have a Sprint Dev unit as well and it had both DJ29 and DK05 pushed to it in pretty rapid succession. Its made the enterprise email (Exchange) unusable so I'm going to set the phone back to DI18 until the problems are sorted out.

i wonder when someone will post DK05. the existing beta has some touchscreen waking issues for me...hint hint

Fastest ROM "on the Market" (2010)

Hey everyone back again with another poll for 2011 results on January 1st, 2011 of what ROM is the fastest ROM period. We ARE NOT talking battery life or anything else just PURE PERFORMANCE. So post your your vote by stating:
- what ROM you are using
- what kernal
- overclock speed
- JIT or No JIT
- and any other things you may have applied to get peak performance from the HTC Hero

Umm Really? In Development?
Let the Flames Begin!!!!

Yea. Definitely wrong section bro. Throw this in the general section.
Does the user have the option to close a thread?

cordell12 said:
Umm Really? In Development?
Let the Flames Begin!!!!
Click to expand...
Click to collapse
Me first!!
You can download ROMs from the Market now?
Mod alerted asking to be moved to General. /sigh.

cordell12 said:
Umm Really? In Development?
Let the Flames Begin!!!!
Click to expand...
Click to collapse
INDEED! *gets out fire extinguisher*

the BEST part of this threads location is that it was actually MOVED here, into development!!! from themes and apps!!!! WTF.

when I get on my laptop later ill move it into general section lol sorry I can not delete the threads on my phone lol

I think you are better off with whatever ROM you have right now... atleast till you learn to read.

il Duce said:
the BEST part of this threads location is that it was actually MOVED here, into development!!! from themes and apps!!!! WTF.
Click to expand...
Click to collapse
Not according to the logs, which say it was originally posted here...

okay after getting the moving and deleting situated its time to get this show on the road....i'll kick it off
after trying damagecontrol, fresh 2.4.0, and liquidsense 2.1 stable my vote goes to.....
ROM - LiquidSense 2.1 Stable
Kernal - Zen's AnyKernel with no Perflock
JIT - N/A
Other Performance Additions - N/A
this by far is the fastest in my opinion....if this rom ends up tying with another rom then i'll go into DEEP explanation as to why i think this is the fasest "on the market" in our debate between the top 2 if there ends up being a tie (btw roms are NOT available on the market i'm just saying that as like the fastest rom available lol )

pulser_g2 said:
Not according to the logs, which say it was originally posted here...
Click to expand...
Click to collapse
sorry if I missed something, just saw it said "moved" in themes and apps.... still, shouldn't be in either, lol.

il Duce said:
sorry if I missed something, just saw it said "moved" in themes and apps.... still, shouldn't be in either, lol.
Click to expand...
Click to collapse
that thread "was never there" in the themes and apps or in the android development sub forum....what are you talking about?
but back to the main topic

I believe Cyanogenmod with Braninfuct kernel is the fastest. Cyanogenmod has much better Graphic drivers than that of any HTC sense rom, which makes the performance much better.

I believe Cyanogenmod with Braninfuct kernel is the fastest. Cyanogenmod has much better Graphic drivers than that of any HTC sense rom, which makes the performance much better.
Click to expand...
Click to collapse
I can respect your opinion so we have two votes now

aospMod is easily the fastest ROM I've ever used. Definitely faster than CM6.1, which I've also used. I don't even need to OC it or anything.

D3FAULT said:
aospMod is easily the fastest ROM I've ever used. Definitely faster than CM6.1, which I've also used. I don't even need to OC it or anything.
Click to expand...
Click to collapse
It's OC'd by default. lol. Every hero rom with a custom kernel is oc'd.

Decad3nce said:
It's OC'd by default. lol. Every hero rom with a custom kernel is oc'd.
Click to expand...
Click to collapse
Correct. I meant not OC'd above what it is by default, which is 691mhz I believe.

I have used a lot of different ROMs and so far the fastest has been NFX stock ROM. CM6.1 was very nice but had moments of lag. I keep a Nandriod of it because I want it be my ROM of choice but when ever I actually you it, I get frustrated and go back to NFX.

I have used a lot of different ROMs and so far the fastest has been NFX stock ROM. CM6.1 was very nice but had moments of lag. I keep a Nandriod of it because I want it be my ROM of choice but when ever I actually you it, I get frustrated and go back to NFX.
Click to expand...
Click to collapse
that is why I love liquid sense b/c IT IS NFX's stock Rom just otimized by NFX and others to make the Rom as fast as possible

-Rom: LiquidSense based off of Nfinite's .7
-Kernel: Decad3nce's #589 w/ SetCPU [email protected]@264
-Mods: Supercircle Battery Mod-&-espresso theme

Quadrant scores for all current ICS roms 3/08/12

I had nothing to do all day at work so I thought it might keep me busy flashing my transformer with all current ICS builds and then running quadrant benchmark. Well, it passed the time anyway.I didn't do any tweaking at all. I just flashed, booted and then ran these benchmarks. Here are the scores if you're interested.
Team EOS nightly build 7
1st score 2604
2nd score 2600
3rd score 2662
average score 2662
Revolution 3.0.1
1st score 2007
2nd score 1984
3rd score 1975
average score 1989
Code Named Android 1.5.0
1st score 2235
2nd score 2372
3rd score 2314
average score 2307
Revolver 4.0.0 Beta 1.3.1
1st score 2396
2nd score 2372
3rd score 2372
average score 2380
CM9 with Cornerstone
1st score 2303
2nd score 2552
3rd score 2465
average score 2440
Spark 20120219
1st score 1854
2nd score 1872
3rd score 1920
average score 1882

Thanks.
What I love about this place is that every question I'm about to ask has already been answered by some enterprising person on here.

But yet you didn't even hit his thanks button.
Thanks are about as important as Quadrant scores but after a post like that it kind of made me laugh a bit.

Can you bench the official CM9 lastest nightly?
Thanks

Thanks are about as important as Quadrant scores but after a post like that it kind of made me laugh a bit.
Click to expand...
Click to collapse
Yeah I agree. There are lots of things not considered here such as kernel and cpu clock speeds and such. I was just bored and thought someone out there might find it interesting to see some "out-of-the-box" (so to speak) scores.
Can you bench the official CM9 lastest nightly?
Thanks
Click to expand...
Click to collapse
Sure, give me a link and I'll give it a whirl.

That wasn't meant for you. I appreciate your initiative in all that entail flashing all these ROMs. It was more for the guy above me showing his undying gratitiude for your efforts, but didn't even have the courtesy to hit your thanks.
Might also want to put a Stock scrore in there too. Just so someone who isn't rooted knows how their stock ROMs compete against custom ROMs. Noobs love Quadrant scores.

great info
Nice job sort of data that is use full to us people to lazy / busy to try out each rom
+1 for stock score

tryfarming said:
Nice job sort of data that is use full to us people to lazy / busy to try out each rom
+1 for stock score
Click to expand...
Click to collapse
+1 for stock too. Thanks!

tryfarming said:
Nice job sort of data that is use full to us people to lazy / busy to try out each rom
+1 for stock score
Click to expand...
Click to collapse
I made a video about this:
http://forum.xda-developers.com/showthread.php?t=1514962

Kafluke said:
Yeah I agree. There are lots of things not considered here such as kernel and cpu clock speeds and such. I was just bored and thought someone out there might find it interesting to see some "out-of-the-box" (so to speak) scores.
Sure, give me a link and I'll give it a whirl.
Click to expand...
Click to collapse
Right here. http://download.cyanogenmod.com/?device=tf101
thanks man!

try Codename android with tweak activated and reboot

Thanks! and please do it for other new ROM that coming

Do this one next:
http://forum.xda-developers.com/showthread.php?t=1542302

Team EOS nightly build 7
1st score 2604
2nd score 2600
3rd score 2662
average score 2662
???
Why average is so big?

Woodrube said:
, but didn't even have the courtesy to hit your thanks.
Click to expand...
Click to collapse
People nagging about not using the Thanks button is just about the lamest thing here on xda.... Even more lamer is people begging for Thanks in their signatures.

Thanks.
What I love about this place is that every question I'm about to ask has already been answered by some enterprising person on here.
Click to expand...
Click to collapse
Woodrube said:
Thanks are about as important as Quadrant scores but after a post like that it kind of made me laugh a bit.
Click to expand...
Click to collapse
Woodrube said:
It was more for the guy above me showing his undying gratitiude for your efforts, but didn't even have the courtesy to hit your thanks
Click to expand...
Click to collapse
Martijn1971 said:
People nagging about not using the Thanks button is just about the lamest thing here on xda.... Even more lamer is people begging for Thanks in their signatures.
Click to expand...
Click to collapse
Wasn't a nag, it was just kind of funny that he expounded so much for OPs efforts but didn't even hit his thanks button. It is about courtesy not numbers.
And I agree about the solicitation of thanks in the signature. When I first came to XDA, I did it but realized that I could actually put something useful in my sig instead.

It's nice, but the best 3D power i have in Codename Android v1.5.0

Martijn1971 said:
People nagging about not using the Thanks button is just about the lamest thing here on xda.... Even more lamer is people begging for Thanks in their signatures.
Click to expand...
Click to collapse
Agree.. You get a thanks for that..
Sent from my Transformer TF101 using Tapatalk

Team EOS Cornerstone Build 4
Test 1: 2602
Test 2: 2811
Test 3: 2812
Test 4: 2808
Average Score: 2758

Stock one for compare if it helps, completely stock with OTA ics
Test 1 - 2342
Test 2 - 2428
Test 3 - 2488
Test 4 - 2470
Average - 2432
I did receive some emails during the test, didnt reboot or anything before hand.
So it looks like from the post above mine the team EOS rom is "about" 13.5% faster.
Decent performance increase.

Linaro-Code will improve Android 4.0.4 !

Hey guys!
In Samsung Galaxy S forum i found something veeeeeeeeeery interesting!
It's about the Linaro-Code that should speed up Android about a noticeable amount!
Just take a look at this video:
Linaro-Video from Youtube
(This is the related article: cnx-software.com)
Idk what i should think about it. If it really works, i guess it would be worth a try ?!
Here is another article, which says, that it can speed up Android 2 times!
"The Linaro team, who’s constantly working on optimizing software for the ARM
architecture, has worked on optimizing the latest version of Android as well, and
they claim up to 2x improvement in performance compared to stock Android
4.0.4, running on the same TI OMAP 4430 hardware. The improvements are
mostly CPU-related, even though they are showing a graphical object to test the
performance. They’ve managed to double the FPS for this particular test, but
they’ve also managed to increase the Sunspider performance by 30%, which is
also quite a big deal."
- quote from this article.

Some one has already posted bout this and its incompatible with our kernel
Sent from my GT-I5800 using XDA

There are there posts in this section
The original started by Persian sphinx and this has already been discuss by hillbeast in his thread
Why isn't any one searching for similar threads
Sent from my superphone using superswift Swype

This is very interesting, but will this work on any 4.0 device? btw, no need to re-post the same thing (no offence)

Jason-EX said:
This is very interesting, but will this work on any 4.0 device? btw, no need to re-post the same thing (no offence)
Click to expand...
Click to collapse
Hillbeast tried to compile it for his serenity builds but got errors along the way i think we need a better kernel to use it
Sent from my GT-I5800 using XDA

It would be compiled in kernel 3.x
Cheers
Sent from my GT-i9003 powered by RemICS S III-looking rom!

bscraze said:
It would be compiled in kernel 3.x
Cheers
Sent from my GT-i9003 powered by RemICS S III-looking rom!
Click to expand...
Click to collapse
well, that will be awesome :good:

Jason-EX said:
well, that will be awesome :good:
Click to expand...
Click to collapse
Yes, it would really be. I just posted it because I thought kernel 3.x could handle it.
Sorry for double-posting this topic.
Anyway; if our developers find more time and are able to make kernel 3.x running
(which would be awesome!) our phone would get an enourmous speed-boost.

We have used it in our CM9 port on LG P500, but there are still some issues to solve, but our Linaro-build is smooth So good luck! If a dev needs some info on how to include it in their code, look at our project-site github.com/lgics