Database Info

Disappointing Linpack Scores

Hey guys, could someone please explain to me why my Nexus S (2.3.3) is only getting an avg score of 12 mflops while my N1 (2.3.3) is getting an avg score of 37 mflops on the Linpack app?

Because benchmarking does not equal real world performance.
Quillcomm Snapdragon processors have better performance when it comes to this particular test.
Hummingbird will excel in other specific benchmark tests.

Run a test with any 3D game or HD, if it lags, then its the device, if it runs smooth, then is the benchmark app itself that hasn't been updated to read scores due to our ota, I had the same thought and issue with quadrant but then I installed modern combat and pes11(soccer game) and both run smoothly, try smartbench 2010 its free on the market and updated for our phones.
Sent from my Nexus S using XDA App

Cool that's good to know. I'll try 3D games and compare. Thanks for the feedback.

lynkx said:
Hey guys, could someone please explain to me why my Nexus S (2.3.3) is only getting an avg score of 12 mflops while my N1 (2.3.3) is getting an avg score of 37 mflops on the Linpack app?
Click to expand...
Click to collapse
linpack is made for the snapdragon processors. my overclocked nexus s gets around 23 on linpack. but, lets see one of those snapdragons do this on quadrant, or get 63 fps on nenamark, or get a 37000+ on an3Dbench...
{
"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"
}

That's a nice score. Well I just d/l Smartbench 2011. My NS got around 850 while N1 got 970, both are estimates. But I did notice that on the corridor test the NS ran it super quick while N1 was sluggish.

Benchmarks depend on what you do to your phone, from a custom rom to an overclocked kernel and other tweaks, such a lagfix do the trick, it looks awesome to hit those numbers but the device cannot be stable 24/7, not for a daily use.
Sent from my Nexus S using XDA App

Quadrant Scores

Hey everybody I'd like to start a thread to show off your T-Bolt Quadrant Scores. I'm using Drod's 1.8 Ghz kernel, and I'm hitting 2500+ consistantly. This phone is wicked fast, and I see it only getting better. So everyone, if you will, follow suit....
{
"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"
}
Sent from my ADR6400L using XDA App

OP, please move this to general since the topic does not have anything to do with dev work. thanks

This does need to be moved.
Tbolt - VirusRom_rc1b1-drod1.8
ghz

What app are you using for the scores? Thx in advance.
Sent from my ADR6400L using XDA Premium App

How do you guys get so high quadrant scores? I get consistently under 2000.

das bamf 1.2 @ 1804/537, performance governor...

Running BAMF 1.3.2 oc 1.2

[URL="http://a.

frnt242 said:
What app are you using for the scores?
Click to expand...
Click to collapse
quadrant standard..

Virus ROM - drod's 1.8 kernel.
sent via r00t3d TB - thx jcase!

not that it matters... lol

CPCookieMan said:
How do you guys get so high quadrant scores? I get consistently under 2000.
Click to expand...
Click to collapse
Overclocked kernals
Tbolt - VirusRom_rc1b2-drod1.8ghz

hey imnever, which app are you using there?

That's Quadrant Advanced. Basically same as standard, lil more detail and no adds.
Tbolt - VirusRom_rc1b2-drod1.8ghz

thanks..i wondered why they called it standard...i dont know if its my rom or imoseyon's tweaks, but i dont have ads so i never put that together.

How's the battery life with that 1.8 rom?

Just threw on the 1.8 kernel for ****s and giggles BLAZIN

Drod 1.8ghz kernel with Chingy's Rom:

Does this mean i win at the quadrant game?

adrynalyne said:
Does this mean i win at the quadrant game?
Click to expand...
Click to collapse
Leave it to Adryn to come and destroy shi*. How the hell did you pull a 3k?
I am luck to even get half that...

{BENCH WARS} / SETTINGS & CONFIGURATION Discussion

Just a thread to post screen shots of your Sensations Bench scores, and discuss what settings, apps, scripts, etc. you're using ,to see how far we can take our dual core and how much we can learn about optimal settings, etc. Lets see who can be the fastest and how they got there. People on Sense Roms feel free to post here too and have your Sense Wars. AOSP ROMs will almost always be higher on bench scores since there is less bogging down the system.
Benches can be manipulated.
For example- turning up the settings/overclocking to run the phone harder and at maximum rates. They may also be manipulated in other fashions. Feel free to question any score.
No SS=No proof.
Use whatever Bench you want. Antutu and CF-Bench are the standards, as Quadrant often gives crazy high or low scores on its own. Though no bench is truly accurate, because when ran at max with Airplane Mode on, they never return the same number. Leaving you to kind of weed out the bad tests and average out your score. They can be used to get a good general idea of performance related settings.
They should also be done at daily driver settings and clocks to find out what it kind of speed we can get out of our DD settings.
Be honest about your settings. This is only for fun and/or seeing what set ups/numbers/clockings work. Which save battery, etc.
{
"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"
}
On CM7 Alpha9.
Running System Tuner Pro maxing both cores at 1.78. Faux CM7 0.2.0 1.78 Kernel.
V6 Supercharger Script set to max ram 100m(setting number 11 or 12)
I like to call this screenshot
"Running with Tabs!"
Sent from where ever I was when I posted this

CF Bench
Sent from where ever I was when I posted this

Sent from my HTC Sensation Z710e

idavid_ said:
Sent from my HTC Sensation Z710e
Click to expand...
Click to collapse
Ha. Already a joker.
Sent from where ever I was when I posted this

sinisin said:
Ha. Already a joker.
Sent from where ever I was when I posted this
Click to expand...
Click to collapse
I just had to post it lol. Seriously though, I get around 2150 on Quadrant, but it's a **** benchmark that can be easily manipulated anyway (see pic above)
Sent from my HTC Sensation Z710e

sinisin said:
Ha. Already a joker.
Sent from where ever I was when I posted this
Click to expand...
Click to collapse
His point is bench marks mean jack sh#t!

idavid_ said:
I just had to post it lol. Seriously though, I get around 2150 on Quadrant, but it's a **** benchmark that can be easily manipulated anyway (see pic above)
Sent from my HTC Sensation Z710e
Click to expand...
Click to collapse
Ill post SS later when I get my new phone in the mail. (Bricked the old one somehow).
1.56Ghz...
My CF bench best is 7040-CM7 build 9, faux kernel (latest)
Quadrant is 4014
Antutu is 6700
Matt

Behold_this said:
His point is bench marks mean jack sh#t!
Click to expand...
Click to collapse
Benches shouldn't be lived by. Quadrant is a very flawed test IMO. On my daily driver settings this is my score. So this is why I say post your settings as well so we can find(hopefully) a good general performance standard for settings/mods/scripts, etc.
Same settings. Just not locking CPU1 online. Also left on max clock.
Sent from where ever I was when I posted this

This is 1188mhz obviously.
V6 Supercharger set at Max Ram with HTK Launcher.
No second core forced on.
Sent from my Sensation using xda premium

Quadrant 2.0

So the new quadrant app is actually somewhat useful. It breaks down the score by category.
Here is what I got the first time I ran it:
{
"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"
}
This is on eclipse 1.3 with pbj kernel and max frequency at 800mhz and conservative governor.
Obviously i have a low cpu score, but somehow i have a massive lead in i/o over nexus s and gnex but not the tab. Do they still use rfs or something? Also everything has a substantial lead in 2d/3d over the gnex, which is strange considering they all have the same gpu and the gnex is clocked faster.
Sent from my pocket-sized, Linux-based computer using electromagnetic radiation... and Tapatalk.

JihadSquad said:
So the new quadrant app is actually somewhat useful. It breaks down the score by category.
Here is what I got the first time I ran it:
This is on eclipse 1.3 with pbj kernel and max frequency at 800mhz and conservative governor.
Obviously i have a low cpu score, but somehow i have a massive lead in i/o over nexus s and gnex but not the tab. Do they still use rfs or something? Also everything has a substantial lead in 2d/3d over the gnex, which is strange considering they all have the same gpu and the gnex is clocked faster.
Sent from my pocket-sized, Linux-based computer using electromagnetic radiation... and Tapatalk.
Click to expand...
Click to collapse
There must be a glitch, or they are not showing the right scores for the other devices. I got scores in the 2100's, better than the galaxy nexus!
Eclipse 1.2 ROM
KAK & Supercharged!

anyone notice the test takes a lot longer now?

bvsbutthd101 said:
anyone notice the test takes a lot longer now?
Click to expand...
Click to collapse
Yeah probably because the test is meant for new phones with 1ghz single core/higher speed dual core processors, not like OG Droid and HTC Hero.

Transformer prime gets around 4000, yikes
Sent from my SCH-I510 using XDA App

I've always found AnTuTu to be a better benchmark. It didn't suffer the seemingly random variability that Quadrant suffered. I'll have to check out the new version of Quadrant to see how it is.

MY QUADRANT SCORE- Droid charge

{
"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"
}
Who thinks they can beat my quadrant score with there Droid charge? Lol. And the only reason I'm posting this here is because I don't have ten post yet.. lol
Sent from my SCH-I510 using XDA App

I remember the days when I thought quadrant meant something. But barely...
There... I helped you get your ten posts
Sent from my SCH-I510 using xda premium

hahahha
it doesnt mean alot... but it does give you an idea of your phones round-a-bout preformance.... for sure..

Like this?
http://www.flickr.com/photos/[email protected]/7018342383/
That was easy. If I'd actually tried, I could have easily broken 2500. That's not a brag...it's more of a commentary on how useless Quadrant is as a performance measure.
Seriously. Quadrant is completely pointless. The number have no bearing on reality. That I/O number in particular is inflated to stupid levels because of the way that Quadrant tests I/O. In Quadrant, the difference between rfs and ext4 is in the hundreds. In the real world, the differences between the two are imperceptible.
Quadrant is the Android equivalent of a penis measuring contest.

SENT FROM THE DOG POUND

stebomurkn420 said:
it doesnt mean alot... but it does give you an idea of your phones round-a-bout preformance.... for sure..
Click to expand...
Click to collapse
I'd argue that too. My original Galaxy Tab running the European GB release scores about 1000 in quadrant. My Charge scores as you saw above. From a user perception, there is virtually no difference between the two. In fact, the Tab edges out the Charge at some things.
I'll say it again. Quadrant is meaningless. Antutu is better. It's still artificial, but the two devices score within 100 of each other on Antutu...that is much more in line with reality.

shrike1978 said:
I'll say it again. Quadrant is meaningless.
Click to expand...
Click to collapse
I agree. It's beyond meaningless. I remember on my Droid2, if I enabled stagefright, I would gain 1200 points on my Quadrant score. It can be faked/cheated too many ways to be useful.

fffuuuuuuckkk...
i am a noob... wat can i say.... hahaha... look like i got some wrk to do...

Ive landed a 3100 a couple times at 1.4
Sent from my SCH-I510 using XDA

What ROM\Kernel are you using?
... and yes I am reply so that I can get my 10 posts........