Database Info

Issue with CPU running bionix 2 w/trinity

i'm running the bionix 2 rom and the compatible ext4 kernel, but when I benchmark it on smartbench 2011, It seems like only 1 core is being used because I get a 1300-1500 productivity score. I flash back to stock, stock bionix, or stock EB and the scores go back to normal (around 3000 productivity). What is going on?

maybe u got a bad download? when dl kernel?

sarni84 said:
maybe u got a bad download? when dl kernel?
Click to expand...
Click to collapse
I downloaded it last night, then redownloaded today, same problem. I actually got normal scores like twice in the beginning, but after that it just seems like only 1 core is being used.

Clear cache and dalvik cache in clockwork and see if that helps
Sent from my LG-P999 using XDA App

aim1126 said:
Clear cache and dalvik cache in clockwork and see if that helps
Sent from my LG-P999 using XDA App
Click to expand...
Click to collapse
also did that each time. I don't think it's just me. People are running quadrant and getting much higher scores(including me). But if you look at the scores for each individual category. You will notice that the only thing that is higher is the I/O due to ext4. the I/O score is so heavily inflated, that the overall quadrant score goes up while some other things are going down. I don't know why everyone relies so heavily on quadrant but it's so inaccurate.

xdmds said:
also did that each time. I don't think it's just me. People are running quadrant and getting much higher scores(including me). But if you look at the scores for each individual category. You will notice that the only thing that is higher is the I/O due to ext4. the I/O score is so heavily inflated, that the overall quadrant score goes up while some other things are going down. I don't know why everyone relies so heavily on quadrant but it's so inaccurate.
Click to expand...
Click to collapse
True that. I hate it when websites run quadrant in their reviews. Specially when the app hasn't even been updated for dual core phones
Sent from my LG-P999 using XDA App

How come some phones get high quadrant scores?

Hey
I overclocked my Galaxy S and I did some tests in quadrant to see if my clock gave me lower performance.. Anyways I got 2950 as my best score @ 1.5 GHz but I see some HD2's getting 4400 in socre @ 1.5 GHz. I thougt my Galaxy S would get a better score then a HD2. Is it becouse the HD2 running a slimmer rom with less background processes or are they just lying?

I believe quadrant scores are heavily varying and dependent on the device, and for the most part, do very little to reflect "real" performance.
I don't have a straight answer, but I wouldn't worry about it too much. Your scores are still magnitude's higher than my Hero, my absolute highest (with unsafe tweaks to the clock and kernel) was about roughly 500

hey guzy....
sorry for the dumb question but...
can u tell ne what overclocking is......??? i'm new to android platform.....
Sent from my GT-S5670 using XDA App

maxcool12345 said:
hey guzy....
sorry for the dumb question but...
can u tell ne what overclocking is......??? i'm new to android platform.....
Sent from my GT-S5670 using XDA App
Click to expand...
Click to collapse
Overclocking is just ramping up the processor in your CPU to higher frequencies for faster performance. There are limits, you can't go too high because you risk damage and overheating, and going too low can cause severe slowdowns and instability.
BUt finding a sweet spot for the minimum and maximum processor speed can result in a much faster phone when the screen is on and a way better battery life with the screen off.

nice.... n thanx.... how did it works.... is there any soft stuff(app) available for it or somethin else....????
Sent from my GT-S5670 using XDA App

LiquidSolstice said:
I believe quadrant scores are heavily varying and dependent on the device, and for the most part, do very little to reflect "real" performance.http://forum.xda-developers.com/newreply.php?do=newreply&p=14894469
I don't have a straight answer, but I wouldn't worry about it too much. Your scores are still magnitude's higher than my Hero, my absolute highest (with unsafe tweaks to the clock and kernel) was about roughly 500
Click to expand...
Click to collapse
Ah...Good old days!
Anyways,there are many many ways to cheat quadrant.Running the app from a ramdisk,patching SQlite,running a rom from SD(I think) all boost the scores.That is,however,because Quadrant relies heavily on I/O,which gives unreal scores.Remove I/O benchmarking from the test and you have a helluva accurate benchmark.Pretty good too.
But on the SGS,let me make my guess.Are you running Cognition rom?If yes,are you using interactive governor?Again,if yes,switch to performance governor and yay,4000+ points in Quadrant!

maxcool12345 said:
nice.... n thanx.... how did it works.... is there any soft stuff(app) available for it or somethin else....????
Sent from my GT-S5670 using XDA App
Click to expand...
Click to collapse
Unfortunately, it requires a custom kernel, which in turn requires a custom recovery (and in some cases, a custom ROM) to achieve.

ohhh crap!!!! thats the reason i want a forum or some dev to create interest in my device..... hey can u tell me somethin j shud do....????
w8 or shud sell my cellphone to buy somethin like optimus??
Sent from my GT-S5670 using XDA App

tolis626 said:
Ah...Good old days!
Anyways,there are many many ways to cheat quadrant.Running the app from a ramdisk,patching SQlite,running a rom from SD(I think) all boost the scores.That is,however,because Quadrant relies heavily on I/O,which gives unreal scores.Remove I/O benchmarking from the test and you have a helluva accurate benchmark.Pretty good too.
But on the SGS,let me make my guess.Are you running Cognition rom?If yes,are you using interactive governor?Again,if yes,switch to performance governor and yay,4000+ points in Quadrant!
Click to expand...
Click to collapse
Im using CM7 with Glitch kernel. And I used "smartass" when I got the score but now I get like 2470 in score now even with performance. Plugged in the charger and now I got 2888 in score so I guess the charger helps a lot in quadrant to..

LiquidSolstice said:
I believe quadrant scores are heavily varying and dependent on the device, and for the most part, do very little to reflect "real" performance.
I don't have a straight answer, but I wouldn't worry about it too much. Your scores are still magnitude's higher than my Hero, my absolute highest (with unsafe tweaks to the clock and kernel) was about roughly 500
Click to expand...
Click to collapse
Cannot agree more. It sure is fun showing around your quadrant score but a quadrant score never reflects true real world performance. Maybe, Neocore would be a better benchmark but then again non of these benchmarks would show how fast apps could load or how smooth you switch between apps.
I use an LG GT540 Optimus. I was able to reach 64.3fps on neocore and 1250 on quadrant after overclocking to some 800MHz with performance controller.
Quadrant also doesn't reflect the core memory/cpu usage. ie Sense UI eats more resources than vanilla. So, same phone running vanilla and sense will have different scores.
I believe Samsung uses TouchWiz. Perhaps, you could score better with a vanilla rom or cyanogenmod.
Last but not the least, Quadrant scores could be faked just as easy as you could fake version numbers in About Phone.

Lirkaren said:
Hey
I overclocked my Galaxy S and I did some tests in quadrant to see if my clock gave me lower performance.. Anyways I got 2950 as my best score @ 1.5 GHz but I see some HD2's getting 4400 in socre @ 1.5 GHz. I thougt my Galaxy S would get a better score then a HD2. Is it becouse the HD2 running a slimmer rom with less background processes or are they just lying?
Click to expand...
Click to collapse
Quadrant sucks use smartbench the sgs is one of the most powerful single core phone out there an smart bench proves it
Sent from my SGH-T959 using XDA Premium App

Sorry for forum!

Jasonhunterx said:
Quadrant sucks use smartbench the sgs is one of the most powerful single core phone out there an smart bench proves it
Sent from my SGH-T959 using XDA Premium App
Click to expand...
Click to collapse
+1 10 characters
Sent from my Inspire 4G using XDA Premium App

to many kernals

hey guys i just need a little bit help here.
im new to the sensation
and i am running Bulletproof 2.2.1
and its very slow, and i even get a ****ty score on quadrant
and i see there are several kernals out there and i wanted to know witch ones are compatible with sense 3.5 bulletproof
and witch one is the best one. by best i mean faster overclocked
thanks in advance

You can try Faux or Bricked, both are great.
Sent from my HTC Sensation 4G using XDA App

Faux 0.2.3 is out soon with a whole host of new features, I've found that I can't get Faux 1.89Ghz version to work (boot loops) but it may work for you. Due to this I use Bricked kernel and then I can overclock (if I want to) up to 1.72Ghz which seems to be the max for my chip!

Here's a really good kernel 2.6.35.14-Bricked-v 1.2-gpuoc 320-ondmnd [email protected] #6 this is what I use and It's awesome so far I've been useing this kernel for three days and It's lighting fast

I'm on BP also and while quadrant is low with stock, I've found stock kernel to be smooth and stable. Today I did flash bricked 1.4, and while my CF scores and quadrant went up significanty, I had lag on homescreens and games like asphalt 6 hd. I have no idea what the problem was. It could have been my Rom, or the way I flashed bricked kernel because I have read so many positives with using bricked.
I went back to stock kernel and all is smooth again, but with low benchmarking scores....
So I guess I'm saying take benchmark scores with a grain of salt. Although it would be nice to have my cake and it too.
Sent from my HTC Sensation Z710e using Tapatalk

whats you quadrant score? i flashed that kernal but i had problems with 3D and he said he fixed it on the latest version 1.4. and when i used setcpu my phone kept rebooting

I don't remember my quadrant score, but CF bench score was really high with bricked 1.4, above GS2.
Now back to stock kernel my CF bench is back to around 5000. Just under the atrix scores.
(But with stock I have no issues with with gameloft games, no lag in asphalt 6 hd.)
I will probably try again in the near future as it could have been user error on my part.
Sent from my HTC Sensation Z710e using Tapatalk

Fk quadrant scores! It all depends on how ur phone is running
Sent from my Sensation using xda premium

g1user101 said:
Fk quadrant scores! It all depends on how ur phone is running
Sent from my Sensation using xda premium
Click to expand...
Click to collapse
I agree!
Sent from my HTC Sensation 4G using Tapatalk

gustav30 said:
I don't remember my quadrant score, but CF bench score was really high with bricked 1.4, above GS2.
Now back to stock kernel my CF bench is back to around 5000. Just under the atrix scores.
(But with stock I have no issues with with gameloft games, no lag in asphalt 6 hd.)
I will probably try again in the near future as it could have been user error on my part.
Sent from my HTC Sensation Z710e using Tapatalk
Click to expand...
Click to collapse
talking about games i havent be able to play them on any rom. some says its not supported and somes dont have the 3d graphics on them

Do the current kernels all have these problems? I would like to know this before I get this phone. And I noticed someone OC'd to 1.8GHz, is this to fix lag or just for fun? Such OC would lead me to question the Sensation's speed quite a bit. Sorry for hijacking the thread, didn't want to make a new one.

Do the current kernels all have these problems? I would like to know this before I get this phone. And I noticed someone OC'd to 1.8GHz, is this to fix lag or just for fun? Such OC would lead me to question the Sensation's speed quite a bit. Sorry for hijacking the thread, didn't want to make a new one.
Click to expand...
Click to collapse
I am running CM7 SelfKang9, with Faux's AOSP 0.2 Kenrnel (1.78ghz OC) and undervolt (Thermal Patch). Everything is smooth as butter and battery life is beyond excellent (getting avg 6mA idle). If you're interested in benchmarks, I have gotten as high as a 6499 AnTuTu score, and a 7424 in CF bench. It does run fine at its rated speed of 1.5ghz, I'm just always gunning for that extra speed.
Edit: For Comparison, my friend has an Atrix, using CM 7.1 and Faux with a 1.3 ghz OC, and he was getting slightly slower benches (6000 AnTuTu). Although both phones run great.

I'm using pyramid3d 8.0.1 along with the bricked 1.2 kernel.
Runs perfect. No lags, and no other problems, although quadrant scores are really low. Quadrant really means nothing IMO.
Sent from my HTC Sensation XE with Beats Audio Z715e using xda premium

pardoned said:
I'm using pyramid3d 8.0.1 along with the bricked 1.2 kernel.
Runs perfect. No lags, and no other problems, although quadrant scores are really low. Quadrant really means nothing IMO.
Sent from my HTC Sensation XE with Beats Audio Z715e using xda premium
Click to expand...
Click to collapse
Sense has a nice bug from HTC that affects Quad & other benches negatively, this can be fixed by using AOSP/CM/MIUI, I noticed this on my previous phones. I think HTC has some lingering debugging enabled in Sense.
Louer Adun said:
I am running CM7 SelfKang9, with Faux's AOSP 0.2 Kenrnel (1.78ghz OC) and undervolt (Thermal Patch). Everything is smooth as butter and battery life is beyond excellent (getting avg 6mA idle). If you're interested in benchmarks, I have gotten as high as a 6499 AnTuTu score, and a 7424 in CF bench. It does run fine at its rated speed of 1.5ghz, I'm just always gunning for that extra speed.
Edit: For Comparison, my friend has an Atrix, using CM 7.1 and Faux with a 1.3 ghz OC, and he was getting slightly slower benches (6000 AnTuTu). Although both phones run great.
Click to expand...
Click to collapse
Thanks for clearing things up for me, I was worried that the phone had speed issues and the OC was to fix that, now I'm not worried with that anymore. Hopefully I order mines next week, but I plan on undervolting & underclocking, just as I did on my MT4G. Although you benches have restored my confidence in the phone.

Lousy Quadrant Score on CNA 1.5.5

Hey guys
I just installed Codename Android 1.5.5 + Matr1x Kernel 18.0 after using CM9 nightly 0304 + Matr1x Kernel 17.5 for a while. I used to run CNA 1.4 but stopped using that as it got very slow after a few hours.
The Quadrant score I was getting on CM9+Matr1x 17.5 was around 2000 with no overclock. Now that I am running CNA 1.5.5 + Matr1x 18.0 i'm only getting a score around 1100...
The ROM feels fine and snappy, but this low score worries me. Is this normal or does this maybe mean there's a problem somewhere?
Thanks!

I'm on same rom same kernel and overclocked to 1430 and only getting 1400s

Ok, so I guess this is normal...
Sent from my Nexus S using XDA

I am running 1.5.5, with Air Kernel 3.6, scary governor and overclocked to 1400, I scored around 2600 with this set up on Quadrant.
I tested the rom after flashing it and it scored it at 1100.
Oh I applied the BM tweaks from CNA utility as well
I am waiting to see how the ram holds up, 1.40 would drain after a few hours, as how the battery lasts, so far it is looking good on both of those items.
I haven't tried to many roms, I liked this one as son as I loaded it, loads all my apps from the market, seems very stable even with it overclocked.
I have read that Gingerbread is really stable and fast but I don't want to loose my native music.flac support on this rom. I have read that 4.04 and 4.05 will have better battery, I am hoping so, as I don't want to go backward to Gingerbread unless I have to.

I applied brain masters tweaks and went from around 1100 to 1800 with no overclock and the rom feels much better now. Now I'll just have to see about battery life but I'm a happy man at the moment
Sent from my Nexus S using XDA

Quadrant score isn't everything. User experience is a lot more important than benchmarks.
Sent from my Nexus S using Tapatalk

You are absolutely right. I use the benchmark simply to measure how the rom is before I get a good feel for it. In this case it feels much better after the tweaks.
Sent from my Nexus S using XDA

ljordan2 said:
Quadrant score isn't everything. User experience is a lot more important than benchmarks.
Sent from my Nexus S using Tapatalk
Click to expand...
Click to collapse
yep .. +1 to that

Clean install using BM rom wipe + Air 3.6 + BM tweaks from CNA1.4 + Deep idle patch for 1.5.5
Deadline + ondemand 100 - 800 with Up Threshold @ 95%.
After 16hrs I've got 68% battery left
Q score between 2000 & 2100 when set to 1000MHz
Sent from my Nexus S using XDA

ljordan2 said:
Quadrant score isn't everything. User experience is a lot more important than benchmarks.
Sent from my Nexus S using Tapatalk
Click to expand...
Click to collapse
I completely agree to this.. I rely on the feel of the ROM when I use them..

ljordan2 said:
Quadrant score isn't everything. User experience is a lot more important than benchmarks.
Sent from my Nexus S using Tapatalk
Click to expand...
Click to collapse
FULL ACK...!

I'm still using 1.4.0 with Air Kernel 3.6 and the supercharger script. I have insanely good performance though battery life has taken a hit over the last few days.
I would like to install BM's tweaks though I'm not sure which version goes with this particular ROM.

If you want higher Quadrant scores try the Trinity T144 kernels. But Quadrant Standard says nothing about the performance of the ROM.. it's just a score.

Yeah, quadrant score isn't everything. The ROM feels fine now but I'm getting the occational freeze requiring a battery pull. I've been running with no overclock but at the moment am running 100/1300 Lulzactive with no issues so far.
Battery life is not good though. I'm at about 50% after 7 hours on battery with light usage

Benchmark scores...

Exactly what do these scores mean along with what is the norm? I have CleanRom 5.5.1 w/ LeanKernel 1.2.1.

Stock JB
Sent from my SCH-I535 using Tapatalk 2

what the scores mean? its running slow?

slufa111 said:
what the scores mean? its running slow?
Click to expand...
Click to collapse
To my understanding it's an estimated test of the processing power/speed of the chipset.

Idk what up with you guys phone but here's mine..
This is stock everything. No root.

Your not that far from me that I should be worried.
Sent from my SCH-I535 using Tapatalk 2

I don't understand how you guys all get such high scores. I've never gone above 4600 on Stock or CM10.......nothing overclocked. Do you guys do some crazy preparation on your phone prior to running tests to get such high scores?

I just flashed the ROM and Kernel and reinstalled all my apps for me...then ran it. If its anything, I kill all apps.

slufa111 said:
I just flashed the ROM and Kernel and reinstalled all my apps for me...then ran it. If its anything, I kill all apps.
Click to expand...
Click to collapse
I did the same
Sent from my SCH-I535 using Tapatalk 2

Very decent battery life.
Sent from my SCH-I535 using xda app-developers app

Benchmarks are really inconsistent in these programs. As long as you think it's fast on YOUR phone, that's all that counts.

Synergy ROM and l535VRBLK3 Modem and the Ktoonsez KT747 Kernel Overclocked to 2106 Mhz
https://www.dropbox.com/s/u9hkv1dozz2fryx/Screenshot_2012-12-17-16-07-56.png

stock unrooted on the jellybean ota update.
6390, getting 6000+ consistently now!

Any running geek bench? I'd like to see those too

chitman17 said:
Synergy ROM and l535VRBLK3 Modem and the Ktoonsez KT747 Kernel Overclocked to 2106 Mhz
https://www.dropbox.com/s/u9hkv1dozz2fryx/Screenshot_2012-12-17-16-07-56.png
Click to expand...
Click to collapse
That's not too impressive for that kind of OC is it? I'm getting 6082 with a completely stock ROM.

I can try it out if you are really interested and post it

The thing with benchmark tests is consistency. Whether it's a PC or a phone, run your benchmark tests under the same conditions or else you'll get unpredictable results. Me, I shut down the phone, pull the battery, reinstall, boot... then wait a few minutes for everything to settle down... then run the test. Background processes will effect your results.
Other than bragging rights, benchmark absolute numbers mean little. They ARE helpful to see what individual changes do...especially helpful after an update or after flashing a new ROM. And use several tests. I have Quadrant Standard, AnTuTu, NenaMark 1 and 2, and Linpack. Also, SunSpider is a great JavaScript test that runs in your browser...meaning you can compare devices running Android, WebOS, iOS, WinX...or even RIM if you're one of the dozen or so people who still use a Blackberry. Lower numbers are better, since the score is in milliseconds to complete the tests.
http://www.webkit.org/perf/sunspider-0.9.1/sunspider-0.9.1/driver.html