Database Users

Hummingbird VS Snapdragon

I cannot understand why everyone is saying that hummingbird processor is better than snapdragon and that's why I started this thread.
I own an HD2 (snapdragon) and SGS (hummingbird).
I've run linpack and quadrant in both phones and here are the results showing that snapdragon is 4 to 5 times faster.
Hummingbird: linpack 13,864 quadrant CPU 1456
Snapdragon: linpack 63,122 quadrant CPU 4122
I'm only talking for the CPU cause if you go to 3D I'll agree that hummingbird is better (but I don't care about 3D cause I don't use my device for games)
Both phones have android 2,2 installed and I have voodoo lagfix installed in SGS

johcos said:
I cannot understand why everyone is saying that hummingbird processor is better than snapdragon and that's why I started this thread.
I own an HD2 (snapdragon) and SGS (hummingbird).
I've run linpack and quadrant in both phones and here are the results showing that snapdragon is 4 to 5 times faster.
Hummingbird: linpack 13,864 quadrant CPU 1456
Snapdragon: linpack 63,122 quadrant CPU 4122
I'm only talking for the CPU cause if you go to 3D I'll agree that hummingbird is better (but I don't care about 3D cause I don't use my device for games)
Both phones have android 2,2 installed and I have voodoo lagfix installed in SGS
Click to expand...
Click to collapse
After looking into it for a while, I was focusing on what makes the Nexus One so much better than the other phones. On the chip level, I didn’t see it. Then it dawned on me to look at what Google had to say on the matter. Well, it was there in black and white. In their 20 May 2010 Developer’s Blog entry (http://android-developers.blogspot.com/2010/05/android-22-and-developers-goodies.html) they say that people could see a 2-5x speed increase. I think it is pointed out in an entry later in the blog dealing with NDK, which I initially missed: “ARM Advanced SIMD (a.k.a. NEON) instruction support The NEON instruction set extension can be used to perform scalar computations on integers and floating points. However, it is an optional CPU feature and will not be supported by all Android ARMv7-A based devices. The NDK includes a tiny library named “cpufeatures” that can be used by native code to test at runtime the features supported by the device’s target CPU.”
So, I guess this means that NEON is the difference. If your phone’s CPU has it and it’s enabled for JIT, you can expect higher Linpack numbers.
Click to expand...
Click to collapse
http://www.greenecomputing.com/2010...ack-scores-so-mucher-higher-than-on-my-phone/
Now stop making topics like this.

the difference you notice is software related

If you want a real test, run a hd video on both phones, or a psx emulator and see if the nexus one is 5x faster... it is the same if not slower then the sgs

Well, SGS got hardware h264 decoding acceleration. Also, maybe you forget, but:
he Hummingbird comes with 32KB each of data and instruction caches, an L2 cache, the size of which can be customized, and an ARM® NEON™ multi-media extension.
Click to expand...
Click to collapse
SAMSUNG and Intrinsity Jointly Develop the World's Fastest ARM® Cortex™-A8 Processor Based Mobile Core in 45 Nanometer Low Power Process
Advanced SIMD (NEON)
The Advanced SIMD extension, marketed as NEON technology, is a combined 64- and 128-bit single instruction multiple data (SIMD) instruction set that provides standardized acceleration for media and signal processing applications. NEON can execute MP3 audio decoding on CPUs running at 10 MHz and can run the GSM AMR (Adaptive Multi-Rate) speech codec at no more than 13 MHz. It features a comprehensive instruction set, separate register files and independent execution hardware. NEON supports 8-, 16-, 32- and 64-bit integer and single-precision (32-bit) floating-point data and operates in SIMD operations for handling audio and video processing as well as graphics and gaming processing. In NEON, the SIMD supports up to 16 operations at the same time. The NEON hardware shares the same floating-point registers as used in VFP.
Click to expand...
Click to collapse
source: wiki
This means Hummingbirds are equipped with NEON. Why its not so effective/used in Quadrant/Linpack? My guess they (these benchmarks) are not compiled/optimised for Hummingbirds, just for Snapdragons.

I came from owning an iPhone and playing lots of games on it. I bought the SGS purely for the gaming performance of the Hummingbird processor.
Having seen the difference in game quality between the HTC Desire and the SGS, I know I made the right decision. Benchmarks don't mean anything.

As long as the device can run apps, games, multimedia smoothly, I dont care much about those benchmarkers, maybe they were designed and/or optimized for snapdragon prior to hummingbird.
Sent from my GT-I9000 using XDA App

i bet you anything he actually doesn't have a sgs...lol
jealousy maybe just a troll, ignore

In terms of overall smoothness (everything, not just games) the SGS is vastly superior to any other android phone I've seen (Desire included).

Darkimmortal said:
everything
Click to expand...
Click to collapse
Really? You have to go all out and use the word "everything" when the phone can get major lockups?
"most things" sounds like a more reasonable and believable choice of words...
Sent from my GT-I9000 using XDA App

My friends I do own an SGS (not happy with it thought) and the tests that I posted were run from me.
I wasn't talking about the gaming performance (I know that SGS is the best out there)
This thread was started so that we can find an answer why is this happening?
I see some answers that cover it but I believe not completely because in everyday use of the phones I see that HD2 is snappier (not much but it is) than SGS (with lagfix).
The best test I believe would be to put the phones to encode something (like a video) but I don't know any software that could do that. (If anyone knows some please point them to me and I'll be happy to post the results here)
The tests you mention with psx and multimedia won't show as what we're looking because the SGS will clearly win because of the GPU.

johcos said:
My friends I do own an SGS (not happy with it thought) and the tests that I posted were run from me.
I wasn't talking about the gaming performance (I know that SGS is the best out there)
This thread was started so that we can find an answer why is this happening?
I see some answers that cover it but I believe not completely because in everyday use of the phones I see that HD2 is snappier (not much but it is) than SGS (with lagfix).
The best test I believe would be to put the phones to encode something (like a video) but I don't know any software that could do that. (If anyone knows some please point them to me and I'll be happy to post the results here)
The tests you mention with psx and multimedia won't show as what we're looking because the SGS will clearly win because of the GPU.
Click to expand...
Click to collapse
man. if you are not happy, then i think you should sell it. no one here will give you a satisfying answer that warm your heart. look for desire hd or something.
to answer ur questions. i get a 2100+ on quadrant. using voodoo fix and oclf on my eclaire. lag free and smooth as butter.
but either way, these test scores mean nothing. they were not designed for samusng hardware. it was designed based on htc and the snapdragon processor.
even people who use neocore for gpu are wrong. if you wana test the gpu performance, use nenamark1. the sgs gives u 49+ fps while the desire HD struggle to give u 35. while if you use neocore. the sgs gives u 56 while desire hd 58
my point is most of those software were designed with htc hardware in mind. so you cant really compare them.
just test your device for your self. apply whatever best roms you find here. if it doesnt lag and smooth for you. then ^^^^ everyone else.
the display alone is worth keepin the sgs for me. sure people might like i phone 4 display more. but nothing in my eyes come close to the contrast and colors of the super amoled. watching a movie or playing a game is a joy in this device.
hell yesterday evening a local htc store had a demo of desire hd. and the guy was nice enough to me play with it for like 1 hour.
device as a hardware look. its friggin sexy as hell. screen ? beauitful large 4.3 screen. quality colors compared to sgs ? fail. a lil slow and laggy " i am sure its because of the firmware. once roms are out, it will be faster "
i was thinking to change to desire hd honestly. but i wake away from the store kissing my sgs.
i love the desire hf look and feel. but as of now its not as smooth as my sgs. and the screen isnt as vibrant.

Psx emulator does not use the gpu...yet
Sent from my GT-I9000 using XDA App

android53 said:
Psx emulator does not use the gpu...yet
Sent from my GT-I9000 using XDA App
Click to expand...
Click to collapse
this. i played king of fighters on my hd2 and it was laggy as hell
smooth as butter on my galaxy s
to be honest. the day psx4droid use gpu. galaxy owners are in heaven.

Its unlikely it ever will though, even modern pc emulators barely use the gpu, only for anti aliasing
Sent from my GT-I9000 using XDA App

johcos said:
My friends I do own an SGS (not happy with it thought) and the tests that I posted were run from me.
I wasn't talking about the gaming performance (I know that SGS is the best out there)
This thread was started so that we can find an answer why is this happening?
I see some answers that cover it but I believe not completely because in everyday use of the phones I see that HD2 is snappier (not much but it is) than SGS (with lagfix).
The best test I believe would be to put the phones to encode something (like a video) but I don't know any software that could do that. (If anyone knows some please point them to me and I'll be happy to post the results here)
The tests you mention with psx and multimedia won't show as what we're looking because the SGS will clearly win because of the GPU.
Click to expand...
Click to collapse
Why in hell woul you want to incodea video using a smartPHONE...?
It's like trying to fit your family and grocery in a sport car... not made for this bro!
stop trying to find reason to "not like" the SGS, if you don't like it, sell it and be done...

Snapdragon/Hummingbird scores in glbenchmark (nexus one/galaxy s):
integer: 20661/27624
float: 11173/7968
I guess glbenchmark uses native C code (hopefully with armv7 optimization), so the JIT compiler has no effect. From the scores it seems that the floating point unit in Snapdragon is faster - but most of the time it is not used (except video & games).
Anyway, a benchmark to measure the same algorithm in both native & java code with scalar & vector instructions would be great...

t1mman said:
Why in hell woul you want to incodea video using a smartPHONE...?
It's like trying to fit your family and grocery in a sport car... not made for this bro!
stop trying to find reason to "not like" the SGS, if you don't like it, sell it and be done...
Click to expand...
Click to collapse
he's not whining, well, not in the first place and i don't see any harm on that i think he's trying to UNDERSTAND reasons behind numbers and daily use with help of other people, so am i. if i had to sell phones for every problem i encounter i will problaby be without (smart)phone at this time
i don't care about benchmarks, but if you think that sgs is smoother than hd2 xda optimized (with wm 6.5 or android 2.2) you obviously never owned an hd2 i'm not talking about games, like johcos says galaxy s performance is not questionable. but android is not all about game. anyway, i don't think hardware is the problem here, sure sgs is superior in many aspects, we know that, regardless benchmarks (even if it seems here that only benchmarks where sgs win are trustworthy, others are not good, not optimized, not realistic, meaningless for real life performance etc.). with a little help from samsung and this community sgs will soon outperform (in real usage) all snapdragon phones. i hope

...when average men talk about the high tech w/o knowledge, boo

ll_l_x_l_ll said:
man. if you are not happy, then i think you should sell it. no one here will give you a satisfying answer that warm your heart. look for desire hd or something.
to answer ur questions. i get a 2100+ on quadrant. using voodoo fix and oclf on my eclaire. lag free and smooth as butter.
but either way, these test scores mean nothing. they were not designed for samusng hardware. it was designed based on htc and the snapdragon processor.
even people who use neocore for gpu are wrong. if you wana test the gpu performance, use nenamark1. the sgs gives u 49+ fps while the desire HD struggle to give u 35. while if you use neocore. the sgs gives u 56 while desire hd 58
my point is most of those software were designed with htc hardware in mind. so you cant really compare them.
just test your device for your self. apply whatever best roms you find here. if it doesnt lag and smooth for you. then ^^^^ everyone else.
the display alone is worth keepin the sgs for me. sure people might like i phone 4 display more. but nothing in my eyes come close to the contrast and colors of the super amoled. watching a movie or playing a game is a joy in this device.
hell yesterday evening a local htc store had a demo of desire hd. and the guy was nice enough to me play with it for like 1 hour.
device as a hardware look. its friggin sexy as hell. screen ? beauitful large 4.3 screen. quality colors compared to sgs ? fail. a lil slow and laggy " i am sure its because of the firmware. once roms are out, it will be faster "
i was thinking to change to desire hd honestly. but i wake away from the store kissing my sgs.
i love the desire hf look and feel. but as of now its not as smooth as my sgs. and the screen isnt as vibrant.
Click to expand...
Click to collapse
Honestly couldn't agree anymore, even with all the problems the SGS has. The screen+hardware combination is just too overwhelming for me to swap the phone for something else.

Quadrant Scores

I'm really curious to see the "Quadrant Advanced" or "Quadrant Professional" scores. In particular, the cpu score. Wondering how 2.3 runs on the Hummingbird, since the Dalvik JIT Compiler in 2.2 didn't really offer the Hummingbird the same amount of cpu performance gain as the Scorpians did.
Can't find it anywhere on the internet, if you get your hands on a Nexus S, please run Quadrant Advanced, and post the screen shot. Thanks!

SamsungVibrant said:
I'm really curious to see the "Quadrant Advanced" or "Quadrant Professional" scores. In particular, the cpu score. Wondering how 2.3 runs on the Hummingbird, since the Dalvik JIT Compiler in 2.2 didn't really offer the Hummingbird the same amount of cpu performance gain as the Scorpians did.
Can't find it anywhere on the internet, if you get your hands on a Nexus S, please run Quadrant Advanced, and post the screen shot. Thanks!
Click to expand...
Click to collapse
Pretty sure there will be plenty of scores on the 16th

slowz3r said:
Pretty sure there will be plenty of scores on the 16th
Click to expand...
Click to collapse
Ya but some people get their hands on it early, like some of the tech sites do, i.e. phonedog. Maybe someone had found a review video showing Quadrant Advanced being run, and could post it Thats all.

SamsungVibrant said:
Ya but some people get their hands on it early, like some of the tech sites do, i.e. phonedog. Maybe someone had found a review video showing Quadrant Advanced being run, and could post it Thats all.
Click to expand...
Click to collapse
I know that if androidandme get ahold of one early theyll bench it

SamsungVibrant said:
I'm really curious to see the "Quadrant Advanced" or "Quadrant Professional" scores. In particular, the cpu score. Wondering how 2.3 runs on the Hummingbird, since the Dalvik JIT Compiler in 2.2 didn't really offer the Hummingbird the same amount of cpu performance gain as the Scorpians did.
Can't find it anywhere on the internet, if you get your hands on a Nexus S, please run Quadrant Advanced, and post the screen shot. Thanks!
Click to expand...
Click to collapse
Quadrant is obsolete. It was designed for Snapdragon architecture.

Engadget just ran the dual core LG Star running 2.2 through quadrant and it only scored 2100. I know Galaxy s phones with the file system fix has beat this easily, which makes me wonder, will the Nexus S have the same file issue problem?
If the dual core lg star was coming out next Thursday as well, I would still get the Nexus S.

the nexus s will have ext4 on the system, data, cache etc. and vfat on the sdcard. so there won't be any file system problem like the SGS already has.

Can't wait for futuremark to release their mobile benchmark and not have to rely on this quadrant bs
http://www.youtube.com/watch?v=tM_3QG4U63I&feature=youtube_gdata_player
Sent from my SGH-T959 using XDA App

Also its been said million times galaxy s lag fix trick quadrant I/O scores not actual performance gain.
I don't see nexus s getting any higher than 16K at moment vibrant around 12-13 with Eugene new non lag fix that's base off new leaked firmware for i9000
Sent from my SGH-T959 using XDA App

I'm fully sure that the Nexus S will have a wonderfully smooth experience.
And I think that this is all that matters.

someone talked about 1631 quadrant score in another thread

bananenlarry said:
someone talked about 1631 quadrant score in another thread
Click to expand...
Click to collapse
Ya I don't know why he made a whole new thread instead of posting it in here. Anyway, I think he referenced phonedog, but I can't find it anywhere on phonedogs site.

SamsungVibrant said:
Ya I don't know why he made a whole new thread instead of posting it in here. Anyway, I think he referenced phonedog, but I can't find it anywhere on phonedogs site.
Click to expand...
Click to collapse
He did it because he was at work at the time, found the info and was excited to share it.

bananenlarry said:
someone talked about 1631 quadrant score in another thread
Click to expand...
Click to collapse
Ummm..... 1631 is not a very compelling score for my N1, so the NS had better exceed 1631 by a mile, or else what's the point???

makelegs said:
Ummm..... 1631 is not a very compelling score for my N1, so the NS had better exceed 1631 by a mile, or else what's the point???
Click to expand...
Click to collapse
Quadrant score is obselete. You will see drastic actual UI improvements.

Anderdroid said:
Quadrant score is obselete. You will see drastic actual UI improvements.
Click to expand...
Click to collapse
How is quadrant obsolete? Mind stating some factual evidence that shows quadrant is obsolete, or were you just stating your opinion as fact?

irishrally said:
If the dual core lg star was coming out next Thursday as well, I would still get the Nexus S.
Click to expand...
Click to collapse
U LIE!!!!!!!

I don't I will still buy NS instead of that LG Star with dual core!
Reasons:
1. I am so tired waiting manufacturer or mobile operator to provide software update. So, pure Google experience is my biggest reason. I want to get the first update, always
Oh yeah, I don't have much time for rooting and ROM flashing.
2. The LG Star dual core benchmark are not that impressive. It is faster, but not by far, not fast enough to be significant. Hummingbird CPU + sgx 540 gpu platform is still not fully utilized.
I think, the dual core is more towards tablet. Good single core platform is more than enough to handle Android mobile phone, at least right now
andyandrwew said:
U LIE!!!!!!!
Click to expand...
Click to collapse

Check it out
YoutubeDotcomSlashwatch?v=wcOMLbIRmoQ
It's only the standard version and doesn't work at all...

Anderdroid said:
Quadrant score is obselete. You will see drastic actual UI improvements.
Click to expand...
Click to collapse
Quadrant, used as a benchmark standard, clearly isn't obsolete, because here we are talking about and comparing quadrant scores.
I think what you mean is that the score itself is pretty much useless as a predictor of user experience. In that sense, I couldn't agree more!!! I've seen ROMs that get higher scores run like total crap, and ROMs with lower scores run like a dream. I've also seen ROMs with high scores run super-fantastic, too!
My point is...that the Nexus S had better outperform my Nexus 1 on Quadrant, otherwise it's just not that impressive of a statement for the Nexus S hardware OR the Gingerbread software, when compared to my N1 (on CM).
I really hope this phone kicks ass, b/c I hope to upgrade my wife's bb to the Nexus S. But, I know that I'm gonna play with a lot, too.... so I want the goods!
Just my .02

My nenamark score, am I crazy?

Yesterday I installed nenamark on my nexus one and ran my first benchmark, and I swear I got 30.1 fps. I specifically noted that it was spot on with movie frame rates.
Every subsequent test I've run I only get around 16fps. Perhaps 30 is way too high for the nexus one's hardware. But did I just imagine my original score or did I actually get that high? What's the highest score we've seen on the nexus one?

Why not just open it again and on the main screen it says what the highest recorded score it...
That'll tell you what your best score it. Also i maxed out on 16.2FPS :-/

Yeah my best run says 17 so I guess I imagined it. But I could have sworn I saw it. Oh well...

It was allllll just a dream ;p

Share your drugs
Sent from my Nexus One using XDA App

42.8 fps nenamark
Galaxy S Captivate

Couldn't using one of these newer kernels with updated GPU drivers be helping?

I got 17.5 fps. Not bad for a year old 1 GHz superphone
Sent from my Nexus One using XDA App

I just got a 22f p s yesterday if you believe it. Same exact kernel and all and it shows up in my high score at least.

My one is 22.3fps, without any tweaks, cleans e.t.c
Miui + wildmonks kernel.

16.4 on the N1 (CM 6.1.1), 36.5 on the galaxy tab
Edit: Remembered I'd underclocked my CPU, ramped it up to 1113Mhz and got 22.9fps. (Which I can't duplicate, can only hit 22.3fps now)

Tell you what, 22 isn't too bad considering this hardware.
My question is how does windows phone 7 devices, that all use the same CPU combo as the nexus one, get silky smooth scrolling thru the entire OS, including very large web pages etc. Do they just use that much better drivers? It definitely shows its possible and the potential.

Smartbench 2011 - now supports multi-core!

Hi guys,
Not sure if any of you are aware but I am the one who developed Smartbench 2010 which is a benchmarking app for Android phones. I've been working on it for the past couple of months and so far, I've been happy with the reception.
But just as it is the case with 99.9% of all apps, it is not capable of utilizing multi-cores.
I am currently developing the version 2011 which IS capable of using multi-cores (as long as Android used is SMP enabled).
This is a BETA version. When it is final, as it was the case with Smartbench 2010, it will be available for free in the Android Market. I am curious to see how well the CPU tests in Smartbench 2011 works with multi-core phones. I am using 4 threads that work in parallel so it should be able to handle up to quad-core processors.
For those who are interested, please download the following and try running it. The results will be automatically submitted to the server and you will see your results in the result chart as well, which is fetched from the server in real-time.
https://market.android.com/details?i...rtbench.eleven
Thanks much!
EDIT: Smartbench 2011 v0.7 has been released. It will now allow you to sort results by either Average, Productivity or Games scores. Please re-install this version.
EDIT2: Smartbench 2011 v0.8 has been released. It now offers a filtering option - you can see all results or just stock clock speed results. Mandelbrot test also has been extended for better precision. Again, please re-install this version.
EDIT3: I have just published Smartbench v1.0 in the Android Market - please install this version moving forward. Results submitted by the BETA version will no longer be stored in the server DB. Your support during the BETA period has been very helpful!

Looks good, my scores went up quite a bit after I closed all the open apps.
Sent from my MB860 using Tapatalk

I think the "multi-core support" code is working pretty good. I can also see that for some scores, they could be more consistent. I think I know why - it is the way I capture the run-time. Instead of getting the average of all runs within the test, I am taking the slowest which might be a bad idea since one thread might have had a bad luck with the scheduling. But other than this, it looks pretty good.
If anyone has any other feedback, please let me know!

Very cool. Nicely done app and there's a big difference from 2010 and 2011 scores.
Sent from my MB860 using XDA App

waltah! said:
Very cool. Nicely done app and there's a big difference from 2010 and 2011 scores.
Sent from my MB860 using XDA App
Click to expand...
Click to collapse
Yes, and you shouldn't compare 2010 and 2011 scores together. The reason I treat them as a separate app altogether is because the apps are tuned differently and the baselines are also different. In 2010, N1 was the baseline, obtaining 1000 across the board. For 2011, G2 was used as the baseline, so N1 will no longer get 1000/1000 scores.
The idea here is to keep up with the technology as it evolves so that the benchmark app doesn't become useless over time. Dual-core was the main focus in 2011 version.

It appears that on Atrix, the "String" test is returning very inconsistent results - probably due to triggering garbage collection as I am using very large strings and creating a lot of them. I am currently working on a fix for this.
EDIT: A new version has been placed in the above location. Please use this version moving forward. Hopefully, this one will report more consistent results for the "String Test". Other tests look pretty good now.

Smartbench 2011 v0.7 has been released. It will now allow you to sort results by either Average, Productivity or Games scores. Please re-install this version.

Holy...my score shot up like no tomorrow.
Well, not the gaming score. It's slightly below the stock one at 2368.
But the productivity score...3040. :O

Heh I hope you don't mind... I ran this on my Nexus S. Only got 800 something productivity and about 2900 gaming, which is to be expected if the baseline is a G2 which is known to do better floating point than the hummingbird but much worse at GPU tasks.
The iPad2 was just announced and it'll most likely have a dual core PowerVR SGX543 which probably means the iPhone 5 will have this same processor/GPU setup. How do you think Tegra 2 will compare to it?

Awesome..getting some great results
Sent from my MB860 using XDA App

i love how they have score of different phone with custom roms

NoNameAtAll said:
Holy...my score shot up like no tomorrow.
Well, not the gaming score. It's slightly below the stock one at 2368.
But the productivity score...3040. :O
Click to expand...
Click to collapse
From what I can see, it's a bit harder to get stable results in multi-core environment. I think it was a good call to go with average scores.

dinan said:
Heh I hope you don't mind... I ran this on my Nexus S. Only got 800 something productivity and about 2900 gaming, which is to be expected if the baseline is a G2 which is known to do better floating point than the hummingbird but much worse at GPU tasks.
Click to expand...
Click to collapse
Actually its helpful to see other results. I just didn't want to have 3000 people bombarding the server at the same time, especially when I am expecting bugs to show up. And yes, I agree with your analysis as well. From what I can tell, Hummingbird based phones are producing results in line with my expectations.
The iPad2 was just announced and it'll most likely have a dual core PowerVR SGX543 which probably means the iPhone 5 will have this same processor/GPU setup. How do you think Tegra 2 will compare to it?
Click to expand...
Click to collapse
Do we know if iPad2 is going to use a single-core SGX543 or multi-core setup? This is feasible with SGX543. Also, I believe SGX543 is supposed to perform better than SGX540 even in a single-core configuration. So I do expect it to outperform Hummingbird in its current generation form, and probably Tegra 2 as well, especially if they go with multi-core GPU config.

lsxrx7 said:
i love how they have score of different phone with custom roms
Click to expand...
Click to collapse
Yeah. My intention was to display all custom ROM configurations. Unfortunately, I don't think there's a single standardized way to read those ROM names... So I am still stuck with this.... Most of the results are still reading "Stock ROM" because it couldn't determine which ROM was used.

Smartbench 2011 v0.8 has been released. It now offers a filtering option - you can see all results or just stock clock speed results. Mandelbrot test also has been extended for better precision. Again, please re-install this version.

Someone here is still running a very old version (v0.2). Please re-download the APK file from the first post and install it.
Thanks.

EDIT3: I have just published Smartbench v1.0 in the Android Market - please install this version moving forward. Results submitted by the BETA version will no longer be stored in the server DB. Your support during the BETA period has been very helpful!

I got 2038 Productivity and 1759 Games. About 600-700 pts less than the Optimus 2x which might have something to do with pushing more pixels.

crea78 said:
I got 2038 Productivity and 1759 Games. About 600-700 pts less than the Optimus 2x which might have something to do with pushing more pixels.
Click to expand...
Click to collapse
There's also a lot going on in your phone. Depending on what was running when all that threads were working, you may end up with some variances. The best indication is to look at the scores displayed by the app, especially those bars that are represented by many test runs.

Hope you don't mind, I'm gonna post a link to this that you can use on your first post to link people to the program:
https://market.android.com/details?id=com.smartbench.eleven
Easier than a QR code on the computer!

[Q] safe gpu and cpu voltages!?

Hello i have been searching but cant find the max safe voltages for the CPU or GPU thanks if you can link me or help

Every device is different, there is no voltage that is perfect for every device.
Sent from my Nexus 7 using xda app-developers app

FitAmp said:
Every device is different, there is no voltage that is perfect for every device.
Sent from my Nexus 7 using xda app-developers app
Click to expand...
Click to collapse
I know that, thanks for responding. I mean the nvidia specific safe voltage chart so i don't fry my device when overclocking like i did before

Nexus7<3 said:
I know that, thanks for responding. I mean the nvidia specific safe voltage chart so i don't fry my device when overclocking like i did before
Click to expand...
Click to collapse
That is called Stock behavior.
What, you think that Nvidia/Google/Asus sat around in a meeting and said to each other,
"Hey, look, we really could run this thing at 200% of our current maximums without any risks at all, but let's not do that because this tablet market isn't very competitive and nobody will care if our performance is crappy compared to other devices. Let's decide to operate it at only one-half or one-third of it's safe capabilities - for no apparent reason whatsoever - just so we can."

Just let everything as it is on default bro
Gesendet von meinem Nexus 7

bftb0 said:
That is called Stock behavior.
What, you think that Nvidia/Google/Asus sat around in a meeting and said to each other,
"Hey, look, we really could run this thing at 200% of our current maximums without any risks at all, but let's not do that because this tablet market isn't very competitive and nobody will care if our performance is crappy compared to other devices. Let's decide to operate it at only one-half or one-third of it's safe capabilities - for no apparent reason whatsoever - just so we can."
Click to expand...
Click to collapse
Dude there is a "/official nividia voltage chart for tegra 3 somewhere all Tue devas talk about one." Thanks for answering. Yeah they could actually look at the new tegra 3 game console it is at 2.3 GHz and also the one in nexus 7 is under clocked to 1.3 to conserve battery life as it is vey stable to 1.5/1.6 ghZ

Tommy-Geenexus said:
Just let everything as it is on default bro
Gesendet von meinem Nexus 7[/QUOte
There is no default for the gpu on m kernel. Over clocking while theoredicly killing your prossesor faster actually can add life to our tablets becuase they will be able to run more recent things when tegra 3 becomes obsolete and we get better games and a newer os in a few major updates. I benchmark in 3dmark a good 3xs stock when in the 600mhz GPU range and 1.7 prossesor clock so i know this will help us run recent stiff. Just wanted to know is there is a nvidia specific """"chart"""" of some kind or manufacturer's guide to safe voltages and clock speed as some dev said" the voltages are high although within nvidia limits" do you know the limits is what I'm asking...
Click to expand...
Click to collapse

bftb0 said:
That is called Stock behavior.
What, you think that Nvidia/Google/Asus sat around in a meeting and said to each other,
"Hey, look, we really could run this thing at 200% of our current maximums without any risks at all, but let's not do that because this tablet market isn't very competitive and nobody will care if our performance is crappy compared to other devices. Let's decide to operate it at only one-half or one-third of it's safe capabilities - for no apparent reason whatsoever - just so we can."
Click to expand...
Click to collapse
http://en.wikipedia.org/wiki/Planned_obsolescence
Not saying this is the case specifically for the N7, but there are many corporations out there who would do this type of thing (*cough* intel *cough*).

The game console nvidia shield has an tegra 4i which is far more powerful and can be clocked very high, theres no tegra 3 game console.
Safe OC goes up to 1.5 GHz, but look at one x+ still stable @1.7 ghz and 520 mhz gpu.
Sent from my Nexus 7 using xda app-developers app

Nexus7<3 said:
Dude there is a "/official nividia voltage chart for tegra 3 somewhere all Tue devas talk about one." Thanks for answering. Yeah they could actually look at the new tegra 3 game console it is at 2.3 GHz and also the one in nexus 7 is under clocked to 1.3 to conserve battery life as it is vey stable to 1.5/1.6 ghZ
Click to expand...
Click to collapse
All semi houses sell their parts in different speed bins, and a price premium is demanded for the high performance bins. So the OEM gets to decide what the performance range will be - constrained by the price target of their design. The fact that a 2.3Ghz Tegra3 speed bin exists does not mean that all parts have that capability.
In any event the "voltage table" you are looking for is published in the stock kernel source code.
The N7, priced @ $200 never was a premium product. A damn fine value for the money, though IMO.
Hard to believe in the planned obsolescence hypothesis (in this case), as it appears that Google is selling the N7 at a loss. Usually planned obsolescence occurs hand-in-hand with a profit motive.

bftb0 said:
All semi houses sell their parts in different speed bins, and a price premium is demanded for the high performance bins. So the OEM gets to decide what the performance range will be - constrained by the price target of their design. The fact that a 2.3Ghz Tegra3 speed bin exists does not mean that all parts have that capability.
In any event the "voltage table" you are looking for is published in the stock kernel source code.
The N7, priced @ $200 never was a premium product. A damn fine value for the money, though IMO.
Hard to believe in the planned obsolescence hypothesis (in this case), as it appears that Google is selling the N7 at a loss. Usually planned obsolescence occurs hand-in-hand with a profit motive.
Click to expand...
Click to collapse
Back to the topic.... Could someone help me find the sourse code for tegra 3 so i know and others how to over volt to achieve certain clock speeDs