Database Develop

Why is SGS Linpack scores so poor?

From the Greene Computing website (accessible from Linpack app), SGS scores range from 8 (Android Eclair 2.1) to 14 (Froyo).
But I see HTC and Motorola Linpack scores (Froyo) ranging from 30s to 40s.
Also does anyone know SGS Quadrant scores (with lagfix)?
Sent from my GT-I9000 using XDA App

Thats 14 PRERELEASE froyo!
But it's theorised it's due to the way SIMD is designed on Hummingbird. Linpack says VERY little about real performance anyway though.

SGS (stock eclair ROM) with OCLF 2.0 gives me a quadrant benchmark score of ~2150, which just about beats every other phone...

andrewluecke said:
Thats 14 PRERELEASE froyo!
But it's theorised it's due to the way SIMD is designed on Hummingbird. Linpack says VERY little about real performance anyway though.
Click to expand...
Click to collapse
Do you mean that Quadrant is closer to real world users experience compared to Linpack?
From published reviews, it does seem that 2D, 3D games (which is computationally intensive) are generally more fluid on SGS than HTCs
So why does Linpack really indicate?
Sent from my GT-I9000 using XDA App

Prasad007 said:
SGS (stock eclair ROM) with OCLF 2.0 gives me a quadrant benchmark score of ~2150, which just about beats every other phone...
Click to expand...
Click to collapse
Gosh, if Eclair with lagfix scores 2150, Froyo should be off the charts! Can anyone share the numbers?
Is performance gains for OCLF similar to Voodoo?
Sent from my GT-I9000 using XDA App

ckkee said:
Gosh, if Eclair with lagfix scores 2150, Froyo should be off the charts! Can anyone share the numbers?
Is performance gains for OCLF similar to Voodoo?
Sent from my GT-I9000 using XDA App
Click to expand...
Click to collapse
bear in mind that the number doesn't tell much. You can have great number but ordinary performance.
many screen capture shared in XDA, this is one of the screen capture on I9000XWJM7 + RyanZa OCLF
http://forum.xda-developers.com/showthread.php?p=7951649#post7951649
{
"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"
}

I believe you have the basic Quadrant ? Well, I have Quadrant Advanced, and the graph shows sections in the bars for each phone bifurcated as CPU, 2D, 3D performance, etc.. With the lagfix, our I/O section for our phones is significantly elongated (due to the filesystem changes). What do I use to take a screenshot ?

One thing to also add to the balance is that the Galaxy S also has the best GPU available for smartphones:
Here is a GPU comparison for some of the leading smartphones:
■Motorola Droid: TI OMAP3430 with PowerVR SGX530 = 7-14 million(?) triangles/sec
■Nexus One: Qualcomm QSD8x50 with Adreno 200 = 22 million triangles/sec
■iPhone 3G S: 600 MHz Cortex-A8 with PowerVR SGX535 = 28 7 million triangles/sec
■Samsung Galaxy S: S5PC110 with PowerVR SGX540 = 90 million triangles/sec
And for comparison a few consoles:
■PS3: 250 million triangles/sec
■Xbox 360: 500 million triangles/sec
Click to expand...
Click to collapse
you also have a dedicated GPU (PowerVR SGX540 GPU) and LinPack is a mathematical only benchmark, so it only test the capacity for the CPU to makes calculation per seconds (MFlops).
Where MFlops is a good indicator, the uses of multimedia applications on modern smartphone is more GPU intensive so, unless you're doing intensive database application on your phone, MFlops are juste an indication.
You can see a comparaison of the full specs here:
http://alienbabeltech.com/main/wp-content/uploads/2010/04/Cellphonehardwarecompari1.png
You can see a real life test here:
http://www.youtube.com/watch?v=dpP5QljEqow&feature=player_embedded#!

It doesn't test mflops though, it simply tests the speed of Dalvik. Similar to what Nvidia did, it is also possible some manufacturers might begin to optimise specifically for benchmarks (and we don't want that)

The Linpack CPU scores are lower than scores from Qualcomm CPUs because the Qualcomm CPUs have higher throughput SIMD FP units. This means they score higher in the linpack scores, but this does not translate to better performance on a day to day basis.
Quadrant scores with OCLF are not correct. Because of the filesystem changes, OCLF just bypasses the I/O stage of Quadrant, and scores the highest possible mark for I/O. Quadrant scores with Voodoo are a more accurate benchmark, because it actually does the benchmark, rather than just bypassing it.

Here we go again.
The Quadrant scores with the lagfixes are largely irrelevant as it does not in all cases test real world performance. To dumb it down, due to the way some lagfixes are implemented it's not actually real disk reads and writes being tested. Doesn't mean real world performance isn't improved by the lagfixes, because it is. The number in the test just doesn't mean anything. The benchmark has some use when comparing different lagfixes to eachother on the same device, but only to say which one is generally faster, not how much faster it is. Then again, when comparing OCLF to Voodoo it is again not comparable.
As for Linpack, the difference in score is due to the "FPU" (SIMD/NEON/VFP) instructions. Snapdragon (Qualcomm) has a better FPU than Hummingbird (Samsung) does. However, (again) it doesn't make that big a real world difference. Before the Snapdragon and Hummingbird devices, FPU instructions were either slow, really really slow, or emulated in software as the hardware for it was simply non-existent in the chips used. The expected performance of tests that use these instructions by Linpack is likely a whole lot lower then is now being reported by Snapdragon, with Dalvik JIT optimizations for this FPU. The part of the total score that can be attributed to the FPU is therefore blown completely out of proportion, as it completely overshadows the performance of the tests that primarily use the CPU.
Of course, yes, Snapdragon's FPU is a whole lot faster than Hummingbird's. The implied real world performance difference by Linpack is however complete nonsense.

To Chainfire, thanks for the detailed explanation on the two tests, and why the Hummingbird and Qualcomm cpus differs in scores.
From anecdotal comments in reviews (see AnandTech review on SGS devices, which I feel is more objective than most reviewers), SGS is generally regarded as the smoothest Android device amongst the current crop of 1st Ghz smart phones. This is largely based on Android UI operations and 2D/3D games performance.
Hence, I was surprised that SGS Linpack scores are so much lower than Qualcomm devices. Your insightful posting has helped to clear that up.
On the topic of SGS performance, lag fixes seem to help tremendously. Is there a compendium introducing the various lag fixes and which may be most suitable for I9000 international devices?
From reading disparate threads, it appears that OCLF came first (using Ext2 file system) followed by Voodoo (Ext 4). From the view of maintaining compatibility with upcoming Froyo and possible future fixes from Samsung (i.e. Compatibility with Kies is a must), which is the better choice?
Note, I am using stock ROM (Eclaire JG4) with ADW.Launcher. and my SGS does not support 3 button recovery mode.
Sent from my GT-I9000 using XDA App

I'm sorry to bump this old thread again, but I felt like it was a better choice than opening a new one.
It makes sense there's a difference between Snapdragons and other cores that are more closely related to the Cortex A8 like the Hummingbird and the OMAPs. But I find one thing weird:
This is a screenshot of an iPad having finished a Linpack benchmark. As you can see it's getting a score of more than 62 MFLOPS. iPhone 4's with a similar Apple A4 (albeit at a lower clock speed) are scoring around 36 MFLOPS, which I confirmed on a friend's iPhone 4 as well as internet sources. Now: the Apple A4 and Hummingbird are supposedly very related, and the biggest difference as I understand is actually the GPU, not the CPU core. So these large differences between an iPad and a Froyo Galaxy S should simply not be there.
To me, this can mean three things:
The Linpacks for iOS and Android are completely incomparable
Samsung and Texas Instruments CPUs can't take as much advantage of the JIT in Android as Snapdragons can
There is a large difference in MFLOPS performance between Snapdragons and Cortex A8's - Snapdragons would get a much higher score even when running iOS
To get a definitive answer about the Galaxy S's comparative MFLOPS performance, I think the best idea is to run a native (not using Android) benchmark on both a Hummingbird and Snapdragon (and maybe an OMAP). Could Ubuntu on a Nexus One and Galaxy S give us a definitive answer? Can anyone test?
This should be helpful for Motorola owners as well.

You can't compare them unless you do a native benchmark on Android.

DCKing said:
I'm sorry to bump this old thread again, but I felt like it was a better choice than opening a new one.
It makes sense there's a difference between Snapdragons and other cores that are more closely related to the Cortex A8 like the Hummingbird and the OMAPs. But I find one thing weird:
This is a screenshot of an iPad having finished a Linpack benchmark. As you can see it's getting a score of more than 62 MFLOPS. iPhone 4's with a similar Apple A4 (albeit at a lower clock speed) are scoring around 36 MFLOPS, which I confirmed on a friend's iPhone 4 as well as internet sources. Now: the Apple A4 and Hummingbird are supposedly very related, and the biggest difference as I understand is actually the GPU, not the CPU core. So these large differences between an iPad and a Froyo Galaxy S should simply not be there.
To me, this can mean three things:
The Linpacks for iOS and Android are completely incomparable
Samsung and Texas Instruments CPUs can't take as much advantage of the JIT in Android as Snapdragons can
There is a large difference in MFLOPS performance between Snapdragons and Cortex A8's - Snapdragons would get a much higher score even when running iOS
To get a definitive answer about the Galaxy S's comparative MFLOPS performance, I think the best idea is to run a native (not using Android) benchmark on both a Hummingbird and Snapdragon (and maybe an OMAP). Could Ubuntu on a Nexus One and Galaxy S give us a definitive answer? Can anyone test?
This should be helpful for Motorola owners as well.
Click to expand...
Click to collapse
You can't really compare the iphone/ipad to android even though the hardware is similar. Android uses a VM so your score is highly dependent on the efficiency of the JIT. This is why you get a much higher linpack score when using 2.2 then 2.1. On a SGS you get around 7-8 MFLOPS with 2.1, and nearly double that 14 MFLOPS if you use 2.2 due to optimization of the JIT. While that's an impressive gain, 2.2 brought more optimization to the snapdragon line of CPU's. Mainly because they have 128 bit SIMD (compared to 64 bit on hummingbird) you get around a 4x increase in performance to around 40 MFLOPS. Someone will surely correct me but the 4x gain on the Snapdragon compared to the 2x gain on the hummingbird is basically because the Froyo JIT is able to send two 64 bit instructions at a time to the 128 bit SIMD in the snapdragon that is why there's a larger gap in linpack scores between the Snapdragon and Hummingbird CPU's in Froyo 2.2 compared to Eclair 2.1.

LeeBear said:
You can't really compare the iphone/ipad to android even though the hardware is similar. Android uses a VM so your score is highly dependent on the efficiency of the JIT. This is why you get a much higher linpack score when using 2.2 then 2.1. On a SGS you get around 7-8 MFLOPS with 2.1, and nearly double that 14 MFLOPS if you use 2.2 due to optimization of the JIT. While that's an impressive gain, 2.2 brought more optimization to the snapdragon line of CPU's. Mainly because they have 128 bit SIMD (compared to 64 bit on hummingbird) you get around a 4x increase in performance to around 40 MFLOPS. Someone will surely correct me but the 4x gain on the Snapdragon compared to the 2x gain on the hummingbird is basically because the Froyo JIT is able to send two 64 bit instructions at a time to the 128 bit SIMD in the snapdragon that is why there's a larger gap in linpack scores between the Snapdragon and Hummingbird CPU's in Froyo 2.2 compared to Eclair 2.1.
Click to expand...
Click to collapse
Scorpion can only issue one arithmetic instruction per cycle, be they 128-bit SIMD or 32/64-bit VFP (scalar). I doubt the JIT is capable of vectorizing data on-the-fly.
I'd attribute the increase FP performance to the fact that Scorpion's FPU is fully pipelined whereas the FPU of the A8 and A9 are not.

You're being excessive.
Because benchmarks mean nothing. Why does it matter? If your performance is good, why benchmark? It's just a placebo. Just go ahead and remove all of your benchmark tools. It's freeing.

[Kernel][AOSP] Tiamat 4.1.0 | 2.6.38.8 | 10/12/2011

​
AOSP Kernels for HTC's 8x50, 7x30, and 8x60 Devices
Also available for the Motorola Xoom​
Tiamat kernels are designed for use on all ROMs that are built from the AOSP source code. This includes ROMs built from MIUI, CyanogenMod, and others.
Tiamat receives no support for use with ROMs based on HTC's Sense - use at your own risk.
Click to expand...
Click to collapse
Tiamat Kernels​​
You can find full details about Tiamat Kernels at our website. The site is up and running and serves as a more centralized location to get updates, downloads, and changelogs for all Tiamat Kernels. There is no forum or Registration, it’s just a more convenient way to keep things organized as we work to add support for more devices.​
Click to expand...
Click to collapse
Support
Join the Tiamat Kernel developers on IRC at irc.freenode.net, #tiamat. Support and questions are generally handled faster there than the forums. You can easily join via webchat here.​
Click to expand...
Click to collapse
​
Special Thanks to:
toastcfh, slayher and the CyanogenMod team for the base kernels and everything else they do for the Android community
bcnice20 for generally being awesome
TeamWin for also generally being awesome
netarchy, chad0989, cuviper, and invisiblek for some great code
intersectRaven and redstar3894 for the Mjolnir compiler
JasonK75 for updating threads​
Click to expand...
Click to collapse
​
Quick Links​
Click to expand...
Click to collapse
8x50 Changelog
8x60 Changelog
7x30 Changelog
Downloads
FAQ
Source Code​

Our site is up.
tiamat-aosp.com
Check it out and give us feedback guys.
Also, join us on irc!
irc.freenode.net
channel - #tiamat
Userbars: -- thanks to -viperboy-
Founder
{
"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"
}
Staff
Contributor
Tester
User

reserved for later

Downloading this now to try out; I assume that you included both Evervolv's and r2doesInc's GB ROMs in your statement of "all current GB ROMs"? That'd be great, because I swear that Invisiblek's kernel, fast and fluid as it is, is causing some majorly excessive overheating in my phone. I'd prefer that my phone break down...well, never, but if it did, closer to my contract upgrade date. lol hoping yours puts out less heat (and I'm only running the main profile at 1.1GHz, so it's not even that much over the stock).

Question: What exactly is HAVS and what is this undervolted to? And I don't know how to test USB tethering, sorry. Also I am running Cyanogen 6.1
I have installed this and ran a quadrant score and linpack test that I will post a screenshot of. I overclocked to 1.15 w/ smartass governor as default:

Going to look at this
Sent from my ADR6300 using XDA App

Cayniarb said:
This is the second iteration of my kernel, but it is the first time I'm posting it on XDA. You can find the post for the first version on the miui-dev forums here.
Click to expand...
Click to collapse
So far no problems with the kernel.
Does your kernel offer any built in screen off profiles when using the Smartass governor? Some kernels with smartass usually do a min-max screen off profile of 245.
Just checking, in case I have to make a screen off profile in setcpu.

dccoh said:
So far no problems with the kernel.
Does your kernel offer any built in screen off profiles when using the Smartass governor? Some kernels with smartass usually do a min-max screen off profile of 245.
Just checking, in case I have to make a screen off profile in setcpu.
Click to expand...
Click to collapse
Second this, i'd like to know about all of this too.

dccoh said:
So far no problems with the kernel.
Does your kernel offer any built in screen off profiles when using the Smartass governor? Some kernels with smartass usually do a min-max screen off profile of 245.
Just checking, in case I have to make a screen off profile in setcpu.
Click to expand...
Click to collapse
Screen off on smartass sets to 128/384 min/max.

morph3k said:
Question: What exactly is HAVS and what is this undervolted to? And I don't know how to test USB tethering, sorry. Also I am running Cyanogen 6.1
I have installed this and ran a quadrant score and linpack test that I will post a screenshot of. I overclocked to 1.15 w/ smartass governor as default:
Click to expand...
Click to collapse
I've seen quadrant scores in the mid 1500's with this kernel. The linpack app always gives kind of low scores since it was updated. With 0xBench, I've seen linpack scores in the 41's.
HAVS is hybrid active voltage scheduling. It allows the CPU to dynamically change the voltage at each frequency (speed) based on need. The ultimate result is better battery life.
This kernel will undervolt as far as 925 mV at lower frequencies. That is a fairly aggressive level and honestly, it may need to be raised a bit. On the other hand, I'm coming from the Evo which is a bit more power hungry (a lot of Evo's can't boot with that low of an undervolt, but my Incredible has not had any issues with it).

Thanks for the kernel. Great to see all the new options popping up from a growing pool of devs. Much appreciated.

Cayniarb said:
Screen off on smartass sets to 128/384 min/max.
Click to expand...
Click to collapse
Thank you for the quick response. Is there a specific reason for the 384 max?
My phone seems to handle the kernel fine, despite it being heavily undervolted. Having it undervolted, yet having a 384 max screen off probably balance out, so im not complaining. I'll continue to test it and see how battery life goes. Thanks again for the kernel.
Happy Holidays!
dc

I am liking the aosp explosion as of late.
Sent from my ADR6300 using XDA App

Cayniarb said:
I've seen quadrant scores in the mid 1500's with this kernel. The linpack app always gives kind of low scores since it was updated. With 0xBench, I've seen linpack scores in the 41's.
HAVS is hybrid active voltage scheduling. It allows the CPU to dynamically change the voltage at each frequency (speed) based on need. The ultimate result is better battery life.
This kernel will undervolt as far as 925 mV at lower frequencies. That is a fairly aggressive level and honestly, it may need to be raised a bit. On the other hand, I'm coming from the Evo which is a bit more power hungry (a lot of Evo's can't boot with that low of an undervolt, but my Incredible has not had any issues with it).
Click to expand...
Click to collapse
I was not trying to say your kernel was bad, in fact 1400's is great for me. I am enjoying your kernel a lot. I just thought you would like some feedback and those 2 things were some quick feedback i thought I could give ya.
DS36 said:
I am liking the aosp explosion as of late.
Sent from my ADR6300 using XDA App
Click to expand...
Click to collapse
Me too. Very nice to see new AOSP ROM's and especially some AOSP Kernel's. Not long ago the choice for AOSP Kernels was very very limited, just kk's and ziggys I think.

dccoh said:
Thank you for the quick response. Is there a specific reason for the 384 max?
My phone seems to handle the kernel fine, despite it being heavily undervolted. Having it undervolted, yet having a 384 max screen off probably balance out, so im not complaining. I'll continue to test it and see how battery life goes. Thanks again for the kernel.
Happy Holidays!
dc
Click to expand...
Click to collapse
I don't like to drop the screen-off max to 245. Maybe this is another hang-up coming from the Evo, but when it is that low, there can be some issues on wake-up. That said, it's a max, so unless the phone needs to clock that high while the screen is off, it wont. Also, even at that level it will still undervolt at either 925 or 950 mV.

morph3k said:
I was not trying to say your kernel was bad, in fact 1400's is great for me. I am enjoying your kernel a lot. I just thought you would like some feedback and those 2 things were some quick feedback i thought I could give ya.
Me too. Very nice to see new AOSP ROM's and especially some AOSP Kernel's. Not long ago the choice for AOSP Kernels was very very limited, just kk's and ziggys I think.
Click to expand...
Click to collapse
I didn't think you were cutting on it, I was just saying that the numbers can go even higher. For my part, this kernel screams. It's not quite as fast as my Evo, but that can handle an OC to 1267, my Inc locks up at anything over 1113. Also, it's probably a bit easier to squeeze some extra performance out of it when cellular data is always turned off.

Only overclocked to 1075. I think I'm in love.
Sent from my ADR6300 using XDA App

I'll give this a try. invisiblek #21 broke my touch capabilities altogether, so we'll see if this one actually works with the 10 point touch.
Edit/Update: My phone does NOT like the 10 point patch, so this kernel is another no-go for me. It figures I can overclock to 1190 but I can't do this, which in the end is probably a lot more useful.

willhill said:
I'll give this a try. invisiblek #21 broke my touch capabilities altogether, so we'll see if this one actually works with the 10 point touch.
Edit/Update: My phone does NOT like the 10 point patch, so this kernel is another no-go for me. It figures I can overclock to 1190 but I can't do this, which in the end is probably a lot more useful.
Click to expand...
Click to collapse
Not every panel is capable of the 10-point multitouch. I can recompile a version that cuts it back down to 3 and that should work for you.
Sent from my PC36100 using Tapatalk

Cayniarb said:
Not every panel is capable of the 10-point multitouch. I can recompile a version that cuts it back down to 3 and that should work for you.
Sent from my PC36100 using Tapatalk
Click to expand...
Click to collapse
thanks for the cayniarb... we miss you in the miui irc gonna load it up now... my panel doesnt like 10 either but why the hell do i need 10 neways... 2 is enough for me
thanks for the great work

[Q] Quadrant Score

hey guys
ive recently flashed this kernel http://forum.xda-developers.com/showthread.php?t=975349
and im getting quadrant scores btw 1400 to 1600
is there any way to increase them
i read somewhere it can be increased by using some file named ' build.prop'

I am on Darky's 10.1 using DarkCore, and I'm getting about 1200 with EXT4 lagfixes enabled.

i used to get 1900 - 2100 back on stock froyo with OCLF 2.1

if it is only about increasing the quadrant score, you could download tegrak overclock tool from market and increase the CPU to 1.25 Ghz..
or, before that you can disable lagfix, switch to CF-root, enable all tweaks (including media stage someting), and then apply overclock and it ll hit around 2350 ...

use damians latest kernel.. quadrant scores of more than 3000 are reported(use jvp as ur firmware) hope this helps..

I just can't understand why people are so obsessed with quadrant scores??? Frankly with stock Rom with ext4 the normal score is around 1500 and with custom ROM you can get around 1700-1800 - that's it (and you can gain around 100 more by removing the journaling of the the ext4 system). Quadrant isn't reliable benchmark and people should stop striving to get higher quadrant scores.
Anyway - if you want high quadrant scores just enable stagefright if you don't mind not being able to play certain video and audio formats on your phone - at least you'll be able to brag about you scores.

quardrant scores shouldnt be an obsession..... but just merely a certain reference once in awhile....

ROM: Miui MCGv 6.13.1
Kernel: Neo7
Quadrant's Score: 1737

JunaidZaka said:
hey guys
ive recently flashed this kernel http://forum.xda-developers.com/showthread.php?t=975349
and im getting quadrant scores btw 1400 to 1600
is there any way to increase them
i read somewhere it can be increased by using some file named ' build.prop'
Click to expand...
Click to collapse
Try F1 JVP ROM as base and go experiment with speedmod kernels i got up to a stable 1.5Ghz with my phone (but very short battery life):
http://forum.xda-developers.com/showthread.php?t=1058814
Also make sure all apps and processes are closed using taskmanager. Then start Quadrant score i got around 2900+ using JVP.

ROM: Miui MCGv 6.13.1
Kernel: Neo7
Quadrant's Score: 1957
No OC but UV between -50 to -125

ROM: XXJVH
Kernel: GTO
Benchmark: 3268
{
"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"
}

Mods need to start closing some of these Quadrant threads, because, there are way too many.
Quadrant scores are great, if you are simply chasing a placebo (or you want the opportunity to grow your ePenis). If you want a phone which is productive though, throw the benchmark in the bin and run a blind test.
You are only testing if Quadrant is faster, not if other applications are. It is possible that optimising for Quadrant may lead to other applications being dramatically slower (because Quadrant isn't detailed enough). You will notice that the Wikipedia article is almost exclusively about the inaccuracies of benchmarking, and that developers generally don't use standard benchmarks (almost every developer will use a test suite that pops out pages of "user unfriendly garbage").
Quadrant simply doesn't provide enough information to tell you what operations is faster.
An optimisation might yield a 6x speedup of directory creation, at the cost of a 2x file writing slowdown. If quadrant is coded poorly and weighs the tests equally, it might tell you the optimisation was worth it (it will be 4x speedup overall). However, in normal usage, you might write to files 10000x more than create directories, so in the real world, it might be actually slower.
It's even less worth it considering that the time most people waste running benchmarks, and optimising their phone (sometimes they spend an hour or so), is never regained in productivity gains again.
That is just an opinion though

Auzy said:
Mods need to start closing some of these Quadrant threads, because, there are way too many.
Quadrant scores are great, if you are simply chasing a placebo (or you want the opportunity to grow your ePenis). If you want a phone which is productive though, throw the benchmark in the bin and run a blind test.
You are only testing if Quadrant is faster, not if other applications are. It is possible that optimising for Quadrant may lead to other applications being dramatically slower (because Quadrant isn't detailed enough). You will notice that the Wikipedia article is almost exclusively about the inaccuracies of benchmarking, and that developers generally don't use standard benchmarks (almost every developer will use a test suite that pops out pages of "user unfriendly garbage").
Quadrant simply doesn't provide enough information to tell you what operations is faster.
An optimisation might yield a 6x speedup of directory creation, at the cost of a 2x file writing slowdown. If quadrant is coded poorly and weighs the tests equally, it might tell you the optimisation was worth it (it will be 4x speedup overall). However, in normal usage, you might write to files 10000x more than create directories, so in the real world, it might be actually slower.
It's even less worth it considering that the time most people waste running benchmarks, and optimising their phone (sometimes they spend an hour or so), is never regained in productivity gains again.
That is just an opinion though
Click to expand...
Click to collapse
Yes, your right and I know it but anyway is something that cans show something of the power of the phone. (sorry by the bad english).
Now, with Miui MCGv6.15 and the Neo7 kernel I've 1939 on Quadrant

Quadrant does not tell anything from speed and you can easily cheat on that test

galaxysdev said:
Quadrant does not tell anything from speed and you can easily cheat on that test
Click to expand...
Click to collapse
Right. But if the dev's or SGS owners don't cheat the test, it could be credible.

You are all alright , but aniway ..
Quadrant : 2843
Antutu Benchmark : 2543
JVP Deodex , Alpha Damian 1 2.2

alasth said:
You are all alright , but aniway ..
Quadrant : 2843
Antutu Benchmark : 2543
JVP Deodex , Alpha Damian 1 2.2
Click to expand...
Click to collapse
Which is your kernel?
[EDIT] I see it now. Sorry ^^

JunaidZaka said:
hey guys
ive recently flashed this kernel http://forum.xda-developers.com/showthread.php?t=975349
and im getting quadrant scores btw 1400 to 1600
is there any way to increase them
i read somewhere it can be increased by using some file named ' build.prop'
Click to expand...
Click to collapse
Okay wouldnt be a better question be whats the fastest rom?
Im still on GingerReal cause it is a combination of both a stable as a fast rom. Quadrant score of GingerReal isnt that high somewhere around 1500 - 1600. I never have forced closes and loading of files etc. go like crazy.
Tried CyanModGen and MIUI which are both great roms but they arent that stable and still have some minor known issues. I'll probably change my ROM until of one of the above have a stable rom.
What is working the best for u guys (and girls offcourse)?
Sent from my GT-I9000 using XDA App

thanks for all your replies guys
i think the better question wouldve been whats the fastest rom in terms of performance?
becuz i do experience a little bit of lag while playing some games

nrbuitenhius said:
Okay wouldnt be a better question be whats the fastest rom?
Im still on GingerReal cause it is a combination of both a stable as a fast rom. Quadrant score of GingerReal isnt that high somewhere around 1500 - 1600. I never have forced closes and loading of files etc. go like crazy.
Tried CyanModGen and MIUI which are both great roms but they arent that stable and still have some minor known issues. I'll probably change my ROM until of one of the above have a stable rom.
What is working the best for u guys (and girls offcourse)?
Sent from my GT-I9000 using XDA App
Click to expand...
Click to collapse
I will try that ROM

How do you OC your device?

I am pretty new to OC so I cant figure out how to do it for 1.5 or 1.6 GHz.
Even when undervolting A1 kernel it is so unstable and cant pass antutu benchmark without crash and reset.
My best values are attached. Its A1 kernel, performance and cfq at 1.4ghz
I hope you will give me advice how to do better cpu benchmark values.
http://oi47.tinypic.com/2ia4ug7.jpg

You need a kernel that supports it. Look for one that is specific to your device.
God promised men that he'd put beautiful women in all corners of the world. Then he laughed and laughed and made the world round

My experience with undervolting is that the gain, battery wise, is nominal, hence I don't care UV.
I remember I got ~7300 in Antutu running CM10 @1500 with interactive governor. But benchmark
isn't everything, and I'm most happy with stock ICS with A1's 1.8.3c @1400 (interactive). I would
prefer running it @1500, but it tends to get a wee bit warm when doing this.

Of course I use kernel designed for my device.
In game overclocking for high performance are numbers important. Every 20pts in antutu benchmark (3d section) does matter in ingame fps (in my case shadowgun deadzone) and you can See it. So nobody has better score? :'(

I believe this is around the highest I got. No, you can't get too much higher. I am quite frustrated at this devices development - the xoom gets much better development and optimisation.... It is all about optimisation.

Nexus 5 Undervolting thread

I found this thread very useful on the N4 forums, so here it is for the Nexus 5.
First, what is undervolting?
No two CPUs are made equally, every one is slightly different. This means they each need a different minimum voltage to run completely stable, so manufacturers set it high enough so that every chip can run without needing a voltage tailored to that specific chip. But this means that most devices can have their voltages lowered, and still run 100% stable
What are the benefits of undervolting?
In my experience, heat reduction is the main benefit. However, this heat reduction has implications elsewhere; for instance it means the thermal throttling will kick in later, so there is less of a performance drop. Also, less power is wasted to heat, meaning marginal gains in battery life. If you cannot run higher frequencies, i.e. over 2.26GHz, undervolting/overvolting these frequencies may help.
Any disadvantages?
Not really - you may experience reboots and instability, but that means you need to raise the voltages. Also, make sure you don't check set on boot until you know the voltages are stable, but if you have any problems you can flash the stock kernel, or in worst case wipe the data from a recovery.
How do I undervolt?
Many CPU apps are available to change the voltages, but personally I use Trickster Mod (donate). Just slowly decrease your voltages, small amounts at a time, and set them on boot when you are happy they are stable.
If you want to really want to get your voltages perfect, you can set the min/max frequency to one you want to modify, and edit each frequency to perfection. To save time, I'd recommend starting with the highly used frequencies - many apps can tell you which your device uses most often.
CPU Binning
I know on the nexus 4 the cpu bin was related to the amount users could undervolt, and Im sure the case is still the same, so heres how you can find out your CPU bin: http://forum.xda-developers.com/showthread.php?t=2515593 It would be useful to include your CPU bin in your post.
Here are my initial voltages, running ElementalX v0.7, CPU bin 3 and PSX - please, feel free to share yours (and your knowledge);
{
"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"
}
Rules of the thread
Be nice to each other, use search if you have a question, and don't flame noobs for no reason.
It would be helpful to include your CPU bin and kernel in your post.
Also, if I have made any errors in the OP or you feel something should be added, quote this (so I get notified) and tell me, I'm sure theres a lot more information I could include!

Reserved

PVS3 -50 Stable

uh60james said:
What's your CPU bin?
Click to expand...
Click to collapse
Just added to the op, I have a 3, which I think is about in the middle? I haven't had a proper look at the thread yet to find out
Sent from my Nexus 5 using Tapatalk

Tom540 said:
I have a 3, which I think is about in the middle?
Sent from my Nexus 5 using Tapatalk
Click to expand...
Click to collapse
I have a 3 too. So far sitting stable at a global -62.5 undervolt the past few days. Tried -75 a few times and got reboots when turning the screen on.

AndrasLOHF said:
I have a 3 too. So far sitting stable at a global -62.5 undervolt the past few days. Tried -75 a few times and got reboots when turning the screen on.
Click to expand...
Click to collapse
Same has just happened to me too, so I'll bump them up to -62.5, although I think I should be able to keep the lower CPU clocks at lower voltages as I'm sure the higher ones are used when turning on the screen
Sent from my Nexus 5 using Tapatalk

I'm running franco r14 and got two almost immediate reboots on your settings. Gonna go a little easier. I have no experience doing this, so it's still a bit confusing.

can probably go a bit easier on the higher clocks, my settings are -75 mv below 1.8 ghz and -50 above. Just played some deer hunter 2014 and seems to be fine. No reboots so farr! My bin is 3 as well. Running stock rom with faux003b6 kernel

I have a bin of 3 as well. I've only been brave enough to go -37.5 on Fauxs latest beta kernel
Sent from the jaws of my Hammerhead!

Got a CPU bin of 2. I went for -50 undervolt for all frequencies on Faux kernel 002 a few days ago, no reboots so far. Haven't tried -62.5 but -75 was too much, so I'll stick with -50.

I was a little bummed when I found out I had bin 2, but I've been running -75mv no problem for the past few days on bricked kernel
Sent from my Nexus 5 using xda app-developers app

You guys are lucky. My phone has bin of 1, could only uv 50mv on high frequency and 75mv on the low end, and I have higher default voltage to begin with.
---------- Post added at 01:33 AM ---------- Previous post was at 01:29 AM ----------
So far I believe the best app to test UV stability is Antutu Benchmark, I could pass StabilityTest no problem but Antutu always FC when I UV bit too hard.

Spunky_Monkey said:
I was a little bummed when I found out I had bin 2, but I've been running -75mv no problem for the past few days on bricked kernel
Sent from my Nexus 5 using xda app-developers app
Click to expand...
Click to collapse
wow.. im have bin 3 and anything past -62 I get reboots... I think I got ripped of on on my cpu

gd6noob said:
wow.. im have bin 3 and anything past -62 I get reboots... I think I got ripped of on on my cpu
Click to expand...
Click to collapse
Exactly the same with me - I think I might be able to set the lower CPU clocks with voltages lower though, I just need a bit more time to test
Sent from my Nexus 5 using Tapatalk

I have a PVS 0 and I can undervolt - 100 without any lockups or reboot. Actually the device hasn't rebooted on me once ever. The only problem is it lacks CPU performance but the GPU works fine.
Sent from my Nexus 5 using Tapatalk

Tom540 said:
Exactly the same with me - I think I might be able to set the lower CPU clocks with voltages lower though, I just need a bit more time to test
Sent from my Nexus 5 using Tapatalk
Click to expand...
Click to collapse
Let me know what your findings are... im just too lazy to test all the configs..

gd6noob said:
wow.. im have bin 3 and anything past -62 I get reboots... I think I got ripped of on on my cpu
Click to expand...
Click to collapse
Bin 3 here, I've been running with - 75 since yesterday without problem... Maybe I'll go lower
Sent from my Nexus 5 using Tapatalk 2

PVS 2 running - 75 volts full speed. Deer hunter runs great so far. Might try - 100 after more testing.
Sent from my Nexus 5 using XDA Premium 4 mobile app

seems like pvs3 is NOT the best for UVing, its all accross the board, im on PVS2 and i think i can do -75 easily , i think im on -50 accross, using franco kernel r14

It might be helpful if everyone included the actual voltages as well, as some have reported starting with different voltages, and some kernels UV put of the box.
PVS 3, I'm running 712.5 @ 300MHz, and 962.5 @ 2.26GHz; check the screenshot in the OP for all voltages
Sent from my Nexus 5 using Tapatalk