Database Develop

[Q] Unleash the real power of Galaxy S SGX 540 GPU

Hello there,
I am just curiuos if there's any way to disable VSync in SGX 540 driver? 3D applications are limited to 56FPS, so Samsung Galaxy S can't win some benchmarks, because of this ridiculous restriction. My SGS always scores about 55.7 FPS or very close to 56, but Droid X has 60 or 62 Hz refresh rate and even though it has weaker GPU (SGX 530) it can still beat SGS in some less demanding benchmarks. The SGS screen refresh rate is 56 Hz... I wish we could unleash the real power of SGX 540, which is now *STRONGLY* limited.
Maybe there are cfg files with options:
VSync=0/1
Or:
MaxRefreshRate 56 Hz etc..
Or something like that... Can somebody look at it?
Thanks in advance for answers!
Best regards,
John

While I'm sure if the framerate was unlocked it would beat every other current Android device in GPU benchmarks, in real-world use this wouldn't mean a thing since the screen itself has a refresh rate of 56Hz...so visually it would be impossible to see any difference other than screen tearing.

It's a limit to save battery, i would like to turn it off/on (for benches) but for real life 56+ fps is just waist of powerbecause of the screen that has 56Hz
(but i prefer 200fps against 60fps on my PC in my 60Hz monitor, somehow i DO notice the difference )

jaapschaap said:
It's a limit to save battery, i would like to turn it off/on (for benches) but for real life 56+ fps is just waist of powerbecause of the screen that has 56Hz
(but i prefer 200fps against 60fps on my PC in my 60Hz monitor, somehow i DO notice the difference )
Click to expand...
Click to collapse
That's called the placebo effect. It's literally impossible for you to see it, you just 'notice' it because you're aware that the framerate is higher, so your mind convinces you that you can see the difference.

AXIS of Reality said:
That's called the placebo effect. It's literally impossible for you to see it, you just 'notice' it because you're aware that the framerate is higher, so your mind convinces you that you can see the difference.
Click to expand...
Click to collapse
Let's try not to prove if human eye can see the difference between 56 or 200 FPS. Let's try to find out if there is a possibility to turn the VSync off to see what SGX 540 is capable of. Because it's an external GPU so it has to have any external panel control or cfg file with some options like NVIDIA and ATI GFX cards have.
Damn I wish I could see 80-100 FPS @ Neocore and GL Benchmark score that kicks ass Droid X score (which is better than SGS, because of damn screen refresh rate)... It's really possible!

About fps perception, that is discussed a lot.... dont bring it here.
The tweak -> would be nice, but just for benchmarks (showing them mine is bigger), or playing with charger connected.
This is mobile device, we got limited power resource.
On pc we got sum competitive games, they provide matchup with players around the world, and it means sth to win/lose, so ppl would do anything to do better.. But in singleplayer...on a 'phone'? No matter to me.

xan said:
About fps perception, that is discussed a lot.... dont bring it here.
The tweak -> would be nice, but just for benchmarks (showing them mine is bigger), or playing with charger connected.
This is mobile device, we got limited power resource.
On pc we got sum competitive games, they provide matchup with players around the world, and it means sth to win/lose, so ppl would do anything to do better.. But in singleplayer...on a 'phone'? No matter to me.
Click to expand...
Click to collapse
Mate! It doesn't matter for the power usage if there are 56 FPS or 200 FPS! GPU is doing the same job with the same power usage but is capped by VSync. It's the software (driver level) limitation, not hardware.

Why don't you just find a more graphic-intensive benchmark instead? It's pointless to turn off Vsync even if we could.
Sent from my GT-I9000 using XDA App

+1 for disable the VSync ....
It`s the reason for the result of GLbenchmark:
{
"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"
}

So what will you guy accomplish from getting higher scores in a synthetic benchmark which doesn't actually represent real speed?
I wish people would stop putting such an emphasis on synthetic benchmarks...

ed10000 said:
Why don't you just find a more graphic-intensive benchmark instead? It's pointless to turn off Vsync even if we could.
Sent from my GT-I9000 using XDA App
Click to expand...
Click to collapse
Testing purposes, to see how fast GPU really is and to kick Droid X ass. ;-) We can't say anything about the real performance of SGX 540... But we can see that every 3D test we run on SGS it gets ~56 FPS...
Mate I can ask why do people overclock their CPUs? Why do people overclock their Galaxy S CPU to 1.2 GHz? It's pointless in some way (battery usage, it's hard to see the difference without benchmark scores etc.). But for power users - people just like me - it's something to test the limits of the hardware.
I am going to check the possibility on the PowerVR forum and keep You updated.
The worst thing is that I can cool my PC hardware using liquid nitrogen but I don't really know mobile software to do the trick.

+1 to try and disable vsync! really would like to know what the gpu is really capable of...

amir_rafie said:
+1 for disable the VSync ....
It`s the reason for the result of GLbenchmark:
Click to expand...
Click to collapse
With tables like this one, for non informed people it may look like Galaxy S is the worst of the current high end devices.

In the /system/build.prop file, there is a line called:
windowsmgr.max_events_per_sec=55
Change it to 60 and see if the framerates go up? Quadrant reports the Galaxy S screen refresh rate is 68Hz.

changed. quadrant gfx test is same 56fps max.

The config file should be somewhere in the OpenGL driver folder I think. I can't check it out because I am at work at the moment...

fua said:
changed. quadrant gfx test is same 56fps max.
Click to expand...
Click to collapse
I changed it to 68 and although the quadrant test didnt go about 57fps, I got higher scores. JPC default got around 950 or so, with this I got about 1000.
The UI is definitely smoother though.

hardcore said:
I changed it to 68 and although the quadrant test didnt go about 57fps, I got higher scores. JPC default got around 950 or so, with this I got about 1000.
The UI is definitely smoother though.
Click to expand...
Click to collapse
Interesting!;-) So we can adjust FPS on the UI right now. Looks like it's possible to make scrolling ultra smooth.

nice, please add all of the findings to post 1 , because we could find it easier later..

amir_rafie said:
nice, please add all of the findings to post 1 , because we could find it easier later..
Click to expand...
Click to collapse
Later. But there are still no idea to walk arount the VSync problem.

Samsung fascinate specs 384 mb ram?

{
"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"
}
What does this mean for us? Possible that we only have 384 megs of ram?

and this is a dev issue how? it belongs in the Q&A section.

Since a lot of people here are trying to figure out how to free up ram up to the 512 limit ( or even close to it) it could belong here.

This has been discussed at length how much ram ourphones have. Posting a picture of a vzw ad does not helped dev's do anything with freeing up our ram.
Sent from my SAMSUNG-SGH-I897 using XDA App

Hmm, just checked gsmarena and there is nothing on the ram of the fascinate, so this might just be true. O.O On gsmarena, it shows as all the other galaxy s phones with 512 mb of ram.
if it is, that kinda sucks, even though you get camera flash, i would gladly take the extra 130mb~ or so of ram :/

Can I just make a quick point here without stepping on anyone's feet?
When you buy a computer advertised to have 4gb of ram, do you actually get to use all 4gb of that ram? No. Your computer takes some of it and divides it up for other purposes. Some is strictly for the OS, some goes and get dedicated to your GPU, and the rest you get to use. The story is the same for our phones. There is a 512mb chip in there for ram. You will only get to use ~380mb of it.
Same for a hard drive. As I've said once before...
Miztaken1312 said:
Sure you buy a 1TB hard drive, but you actually only get to use around 850GB of it due to formatting and whatnot.
Click to expand...
Click to collapse
It's all there, you just won't ever get to use it all for yourself.

miztaken1312 said:
Can I just make a quick point here without stepping on anyone's feet?
When you buy a computer advertised to have 4gb of ram, do you actually get to use all 4gb of that ram? No. Your computer takes some of it and divides it up for other purposes. Some is strictly for the OS, some goes and get dedicated to your GPU, and the rest you get to use. The story is the same for our phones. There is a 512mb chip in there for ram. You will only get to use ~380mb of it.
Same for a hard drive. As I've said once before...
It's all there, you just won't ever get to use it all for yourself.
Click to expand...
Click to collapse
That is cuz how microsoft's file system, NTFS, which counts 1024 kb = 1 mb. That is why it is off. This has nothing to do with NTFS. This is most likely that the ram isnt there. If what you said is the case, that would mean all of the galaxy s phones would be advertised as 304 mb, cuz that is all the is usable at stock.
And your hard drive comparison, thats because for 1 TB to show up in windows, it would have to be 1024 GB instead of the normal 1000 GB just because NTFS has a different way of counting bytes.

miztaken1312 said:
Can I just make a quick point here without stepping on anyone's feet?
When you buy a computer advertised to have 4gb of ram, do you actually get to use all 4gb of that ram? No. Your computer takes some of it and divides it up for other purposes. Some is strictly for the OS, some goes and get dedicated to your GPU, and the rest you get to use. The story is the same for our phones. There is a 512mb chip in there for ram. You will only get to use ~380mb of it.
Same for a hard drive. As I've said once before...
It's all there, you just won't ever get to use it all for yourself.
Click to expand...
Click to collapse
No, that is a really terrible analogy. You buy a 1TB hard drive. You get 1TB, counted in decimal. Just because you don't comprehend the difference between binary and decimal doesn't mean you're suddenly losing space. IT IS THE SAME AMOUNT OF SPACE.
Pretend I point to a can of paint and say it is navy blue. Some other guy comes along and says it is actually 000080 in RGB Hex value. You return your can of pain to the store and yell at the manager that you didn't get navy blue.
That's pretty much what you're saying about hard drives. You are not losing anything at all. Period. End of discussion. A computer simply interprets bits in a different numbering system than what we are used to (binary vs. decimal). There is no magic. There is no loss.
Likewise, your 4GB of RAM analogy is also incredibly flawed. When I buy 4GB of RAM, do I get 4GB? Yes, absolutely. There is exactly 4096KB of RAM there. I can have it sitting on my desk and there is 4GB there. I can install it in my machine and there is 4GB there. What it is used for is irrelevant. It is still 4GB that I paid for. I know that for a fact.
For our phones, we do not know for sure that there is exactly 512MB of RAM and, even if we did, we do not know if there is exactly 208MB of RAM being used for system processes or dedicated to the GPU. It could just be reserved for absolutely no purpose or it could not exist at all (after all, 208MB of RAM is an absurdly large amount of RAM to use for the system when we're still using EVEN MORE for android processes we can clearly see in any decent task manager). You don't know and neither do I.

It has nothing to do with the NTFS file system. It has to do with how the OS counts (megabyte vs mebibyte - base 10 vs base 2).
In regards to RAM, PC manufacturers do state if the GPU is using shared memory. Other than a non-discrete GPU, RAM in a computer isn't typically statically reserved.
The Galaxy S has never had 384~ MB available to the user. The most that can be squeezed out at present, and still have a fully functional phone, is 340~ MB.

So perhaps a developer has more to say about this... since they DEVELOP? It is awful lot of RAM to be used for unknown stuff

Eh, he still has a point. There's 512 there, just some of it is used for the system. I forgive Samsung for this since all android phones (heck, all OS's) suffer this phenomena.
I read somewhere that Samsung chose to statically allocate it rather than dynamically, just to really increase stability. So the cappy may be worse but still you always see this in some form
Sent from my SAMSUNG-SGH-I897 using XDA App

Then why Galaxy Tab shows 444 of available RAM?

Because that's a tablet?
Sent from my SAMSUNG-SGH-I897 using XDA App

The specs of the Galaxy S and Galaxy Tab are the same, apart from the screen resolution.. Even the Quadrant scores are the same.

DirtySoul said:
The specs of the Galaxy S and Galaxy Tab are the same, apart from the screen resolution.. Even the Quadrant scores are the same.
Click to expand...
Click to collapse
You just said the RAM was different.
Anyways, that doesn't change my point that some RAM is always taken by the system OS. Android is a little unique in that each hardware distributor is allowed to tweak the OS so the level of system RAM reserved changes but its still there always
Sent from my SAMSUNG-SGH-I897 using XDA App

fatttire said:
You just said the RAM was different.
Anyways, that doesn't change my point that some RAM is always taken by the system OS. Android is a little unique in that each hardware distributor is allowed to tweak the OS so the level of system RAM reserved changes but its still there always
Sent from my SAMSUNG-SGH-I897 using XDA App
Click to expand...
Click to collapse
They said the amount of *AVAILABLE* RAM shows differently. Insinuating that when two pieces of hardware that are identical show different amounts of available RAM (available, not physically present) that there is something else going on that we the user aren't aware of.
Unknowns like this lead to people asking questions. People asking questions lead to answers. Answers lead to optimizations by brilliant XDA developers. Brilliant XDA developers making optimizations lead to terribly happy users like myself.
So it's in the interest of everyone posting in these forums to ask questions and try to come up with possible answers besides "No".
Yes, we know OSes use RAM and allocates it to various things. I think the issue as stated by others already is that we don't know how MUCH RAM is being used by the OS in our case.
If one device says we have 340 MB RAM available, and another almost identical device says 444 MB RAM available then that means it's POSSIBLE that our brilliant XDA developers could discover how to free up another 100 MB RAM to do fascinating things with.
It's NOT possible for them to discover that if people asking questions and helpful suggestions are shut down by negative people simply stating "NO".

https://github.com/cmsgs/android_device_samsung_galaxys/blob/master/config
you can go here and check out the reserved memory allocation. See line 766 and 1394 under "reserved memory config." the reserved mem adds up to approx. 140 mb. give or take a few. That would put us pretty close to 512. Some of that can be freed up as has been done on the i9000 forum. They have successfully freed up 40-50 mb to be used as free ram, although freeing up too much from certain locations can have negative results such as hindering 720p playback. See speedmod froyo in the i9000 forum for details. [KERNEL] SpeedMod Froyo 340MB RAM Optimized + Sharpness fix by Hardcore. I also think sztupy, developer of z4mod (I believe) had something to do with it.
Just a matter of time and hopefully we will have the same.

I know they're both advertised as having 512MB, but the Tab has more physical RAM than the phone. You can find a post somewhere in the forum talking about the memory chip configuration. The Tab has 128MB more than the phone. That's the primary reason the Tab has so much more available memory.

crazililazn said:
That is cuz how microsoft's file system, NTFS, which counts 1024 kb = 1 mb. That is why it is off. This has nothing to do with NTFS. This is most likely that the ram isnt there. If what you said is the case, that would mean all of the galaxy s phones would be advertised as 304 mb, cuz that is all the is usable at stock.
And your hard drive comparison, thats because for 1 TB to show up in windows, it would have to be 1024 GB instead of the normal 1000 GB just because NTFS has a different way of counting bytes.
Click to expand...
Click to collapse
I just had to chime in - to satisfy the MS/HDD argument, an even better analogy: You purchase a Dell computer that is advertised to have a 500gb HDD. It arrives at your door only for you to find that (GASP!) the Windows install took 10gb, leaving you with only 490gb.
I believe that most if not all mobile handset manufacturers do this to some extemt. My Droid Incredible had 16gb of onboard storage advertised. They didn't mention that 4gb of that was reserved for the actual Android OS. It honestly never crossed my mind to feel swindled - it's just how it is.
That being said, I grow tired of this thread popping up as bumped when I check the Dev Forums for updates. While this discussion certainly has merit, it's nothing new to developers, nor does it have merit for them in particular.
Mods, can we get this thread locked or moved?

Z4mod is from Z4ziggy
tiger4j said:
https://github.com/cmsgs/android_device_samsung_galaxys/blob/master/config
you can go here and check out the reserved memory allocation. See line 766 and 1394 under "reserved memory config." the reserved mem adds up to approx. 140 mb. give or take a few. That would put us pretty close to 512. Some of that can be freed up as has been done on the i9000 forum. They have successfully freed up 40-50 mb to be used as free ram, although freeing up too much from certain locations can have negative results such as hindering 720p playback. See speedmod froyo in the i9000 forum for details. [KERNEL] SpeedMod Froyo 340MB RAM Optimized + Sharpness fix by Hardcore. I also think sztupy, developer of z4mod (I believe) had something to do with it.
Just a matter of time and hopefully we will have the same.
Click to expand...
Click to collapse

[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

[Q] What is and how works VM heap?

Hi
Everybody talk about changing the VM heap value, but... what is the difference after changing it? What about increase its value? And decrease it?
On the CM7 thread adamz667 post this: http://forum.xda-developers.com/showpost.php?p=15065093&postcount=4569
Anyone can confirm that?
Thanks

Increasing cm heap means that apps can use more memory which makes them run faster but then since tbey use up more mem u will have less free memory.
Sent from my Milestone using XDA App

Here you got some definition:
VM Heap Size
The heap is an area of dynamic memory allocated to programs. With a larger heap size, there is more memory for the program and the garbage collector can run more efficiently. Therefore, it is possible to get better performance by setting a larger VM heap size.
However, if you set it too high, you can run out of free memory which can trigger swapping which will degrade performance. Therefore the VM setting is a trade-off.
The optimal value will depend on the RAM on the device as well as the programs and usage patterns, therefore the optimal setting can often only be determined by the user experimentally.
Default value: varies between devices
Click to expand...
Click to collapse
from: http://droidgram.com/forum/8-Android-Dalvik-VM-Heap-Size

adamz667 said:
Here you got some definition:
from: http://droidgram.com/forum/8-Android-Dalvik-VM-Heap-Size
Click to expand...
Click to collapse
Thanks adam, but your post on CM thread was a good theory and this definition it's opposite to your post (a this new one: http://forum.xda-developers.com/showpost.php?p=15092265&postcount=4582)
I only want to know what value use without have to try all of them during days of testing...
Thanks

parads said:
Thanks adam, but your post on CM thread was a good theory and this definition it's opposite to your post (a this new one: http://forum.xda-developers.com/showpost.php?p=15092265&postcount=4582)
I only want to know what value use without have to try all of them during days of testing...
Thanks
Click to expand...
Click to collapse
I have tested all of them today and the best results was with 16mb.
How I was testing it?
Shadow Guardian HD is a very heavy game for MM's ram memory. So I've tested every vm heap size on one and the same level of SGHD, with the same run environment and game was most playable on 16mb (no freezes, reboots, no returns to home screen).

Really 16mb. Which rom? And how far did you get in the game?
Sent from my Milestone using XDA App

thre3aces said:
Really 16mb. Which rom? And how far did you get in the game?
Sent from my Milestone using XDA App
Click to expand...
Click to collapse
I am using CM7rc10 (14.06.2011) I am now in 4th episode (level?) - Medusa Sanctuary.

That is very very interesting. I'm gonna have to give cm7 another chance. Maybe rc10 is a lot better than rc4 (last cm7 I tried)
Sent from my Milestone using XDA App

i also have good performance with 16m tried most settings from extremely large to extremely small...
also png optimized most apps again and again... (which decreased the size of the apps by almost 10mb again - quite sure that it does not make much of a difference)
don't have any app locked in memory and only one homescreen with adw ex without widgets...
dunno till now i have not experienced any issues with lost sms or phone calls...
but i'm aware of that it could happen with my settings so it needs more testing, jit turned off etc. phone works like a charm since i switched from froyomod a week ago or so.
sent from my current location

Are you sure u tried 16 mb. Because I just tried shadow guardian on cm7 rc10 with ur settings and zepplinrox supercharger on balance and tbh it was rather language at times
Sent from my Milestone using XDA App

I've tried jit off and VM heat on 16MB but not get any result.
The launcher was freezing 3 or 4 times, and all games get a minimal improve: 3 seconds more until FC.
I'm testing with: Asphalt 6 and Brothers In Arms 2 (Gameloft games) and Age of Wind v2.2. All versions for MM, with people that play them on stock rom.

thre3aces said:
Are you sure u tried 16 mb. Because I just tried shadow guardian on cm7 rc10 with ur settings and zepplinrox supercharger on balance and tbh it was rather language at times
Sent from my Milestone using XDA App
Click to expand...
Click to collapse
well as i said i do not use that memory intensive apps.
(i think my biggest apps are maps and the browser, and the browser barely goes beyond 30mb, so 2 vm's with 16mb, instead of 2 with 24. but most apps stay below 10mb so i think for my use it is a good solution)
also i debloated the rom massively and replaced most system apps with smaller ones.
cm7 performs very well with this settings, at least for me. (no redraws, no lags in ui no freezes etc...)
like i said i don't use games on my stone
edit: only genie news & weather app reloads often but i think it simply has a high priority.
sent from my current location

That is quite interesting. On ho no I
And 20mb with balanced 3 supercharger works great. But not so well on 16mb
Sent from my Milestone using XDA App

I had tried 16mb weeks ago when somebody suggested it.
That lasted about 5 minutes.
Scrolling between adw screens was brutally choppy.
Swipe finger, wait, look at watch, then bam - the screen suddenly decides to quickly jump to the next screen...

zeppelinrox said:
I had tried 16mb weeks ago when somebody suggested it.
That lasted about 5 minutes.
Scrolling between adw screens was brutally choppy.
Swipe finger, wait, look at watch, then bam - the screen suddenly decides to quickly jump to the next screen...
Click to expand...
Click to collapse
I've got the same with 12mb vm heap. But when I use 16mb everything works fine on CM7 and HO!NO! CM7. Maybe it depends on rom/apps you use and your own configuration.

I guess.
But I'm used to fast, smooth and lots of free ram no matter what rom I'm using
Sent from my Milestone using Tapatalk

adamz667 said:
I've got the same with 12mb vm heap. But when I use 16mb everything works fine on CM7 and HO!NO! CM7. Maybe it depends on rom/apps you use and your own configuration.
Click to expand...
Click to collapse
yeah 12mb is to small... experienced some errors with it too.
i have to say that for me everyhing between 16 and 32 mb seems to do not do much of a difference, but with 32mb my system uses some more RAM right after boot.
after a while it usally uses less (around 80-90mb sometimes less, sometimes more) so i normally have around 60mb free RAM, which seems to be enough.
i'm out of this discussion for now, but i have to say that i rly think ur VM Heap settings should match the apps u use.
so just try different settings.
@Zeppe
yay never experienced that, but i use only one Homescreen anyway. and ADWEX is rly memory friendly : )
sent from my current location

Juicing Speed Out of This Phone

There does not seem to be a thread on what we can do to get the most speed out of the phone, only individual posts in the different ROM threads, now buried in the posts of times. So here one is, even if it is in the wrong thread. As always, your mileage may vary.
===================================
Tweaks Built Into CM9/CM10
===================================
These are basic tweaks built into CM9 and CM10, in the settings app. All ROMs based off of these should also have these settings.
{
"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"
}
The first one is just to make it feel faster.
Settings >> Developer Options
Changing these three values will change the speed of the animation. Change them all to ".5" to make things feel faster.
NOTE: Changing "Window Animation Scale" to anything that is not 1x in Camcory's CM10 will disable the CRT Off lock animation.
Settings >> Performance
16 bit transparency
I have gotten visual artifacts when using it, so I haven't ever used it and the goal is speed and stability.
Surface Improvement
Run wild. None of these seem to make a performance or stability difference.
Processor
CPU Governor: The two best options (in my own opinion) are "OnDemand" and "Conservative." They only difference between them is how fast the ramp up your CPU frequency. Conservative is slower, and gives slightly better battery life.
We are not overclocking today, since we are trying to remain as stable as possible, and because SmartGuy's OC Kernel has disappeared.
I/O scheduler
We're not going there yet.
Memory management
zRam. Debatable, but 10% or 18% seems to work fine (with no adverse affects.)
The others: absolutely turn on.
===================================
Pimp My Rom
===================================
Pimp My Rom. The All-in-One toolbox. It is hard to go into detail, and none of them are "bad." I'll just tell you what I stay away from, either because we cover them elsewhere, or I have no idea what they do.
Before you try anything here, "Tools >> Rescue Package" to download the flashable zip to fix PmR problems.
Things aren't supposed to go wrong, but they do sometimes. It's a failsafe.
Now, I don't touch:
Tweaks >>Multitasking (gets covered later)
>>Dalvik VM (no idea what it does)
>>Miscellaneous >> "Optimize sqlite3" and "Zipalign all..." (get covered later)
===================================
The V6 Supercharger and KAKernalizer
===================================
Script Manager is needed to run V6 and KAK.
Place both scripts that you downloaded from the fourm page onto your phone. Anywhere accessible. Find 'em and Run 'em. Make sure the "Su icon is pressed"
On KAK, just run the first option.
On V6, install everything, including the Nitro Lag Nullifier (option 21 I believe) and run option 12. Make sure the Min Free values match.
You may have to re-run option 12 after a reboot. And if your phone is lagging, try running option 17, the Engine Flush, to free up some resources. I have this one set up to run hourly with Tasker.
Hey we just got the "at boots" and the OOM killer settings from PmR covered and the I/O Scheduler from the Settings covered at the same time!
===================================
Seeder
===================================
Seeder is relatively new to the speed scene, and has already raised and uproar receiving comments for and against from Cyanogen and Google Android Devs alike. Whether it is a real lag killer or a placebo, I've noticed a difference, and that is what counts, right? how it feels.
This one is simple. Download the apk, install, open it up, enable and set at boot.
===================================
ROM Toolbox
===================================
This app hails from the makes of "SuperUser" and "Busybox Installer," the essential apps in any rooting process. Now here we will be culling our start-up apps only, but there are many other feature you can goof around with. Just be careful out there son.:crying:
Tools >> AutoStart Manager"
Now you can disable the apps you think you should disable.
Feel free to go into the other events and disable apps, but be warned: some apps will cease to work if their call signs are disabled.
As far as other things tips go, just be careful apps you install. The kindof app that run all the time are the ones you want to download very infreqently. I also use Watchdog to see if any apps are eating all my CPU and use Rescan Media Root to stop "Android Media Process" when it runs amuck.
Of course, there is only so much you can do. Even with these, you will still have a phone that lags and can't run all the best games. Hopefully this just makes it more bearable for you.
Feedback is appreciated. Comment on what works and what doesn't, as well as other things you've tried to make things run smoothly! Happy hacking!
​

Just tried seeder. I think it made a difference?
Sent from my SGH-T679 using xda app-developers app

I'm working on linaro cm10, on full linaro the camera doesn't work, so I have a ROM with everything strict alising, and linaro optimesed except the lib folder, which only takes 100 points or so off the benchmark.
Sent from my SGH-T679 using xda app-developers app

I think the only issue with all these tweaks is that there is too much of a good thing if you will. That's been a major downside around the forum here. The one thing I have learned is that you can tweak this phone from here to eternity but you will never get that "SIII" performance people seem to think they will get. You got a single core processor and 512mb of RAM so it is what it is. Don't get me wrong, a FEW tweaks are helpful but if you overdo it they will all cancel out each other.
Sent from my SGH-T679 using xda app-developers app

ArtfulDodger said:
I think the only issue with all these tweaks is that there is too much of a good thing if you will. That's been a major downside around the forum here. The one thing I have learned is that you can tweak this phone from here to eternity but you will never get that "SIII" performance people seem to think they will get. You got a single core processor and 512mb of RAM so it is what it is. Don't get me wrong, a FEW tweaks are helpful but if you overdo it they will all cancel out each other.
Sent from my SGH-T679 using xda app-developers app
Click to expand...
Click to collapse
Yeah, I tried playing GTA vice city, it laggs so bad even over clocked at 1.9ghz
Sent from my SGH-T679 using xda app-developers app

I've noticed a pretty big difference with seeder. Im about to flash smartguys3.0 kernel and see how blazing it runs with that.
IRX120 said:
Yeah, I tried playing GTA vice city, it laggs so bad even over clocked at 1.9ghz
Sent from my SGH-T679 using xda app-developers app
Click to expand...
Click to collapse

ArtfulDodger said:
I think the only issue with all these tweaks is that there is too much of a good thing if you will. That's been a major downside around the forum here. The one thing I have learned is that you can tweak this phone from here to eternity but you will never get that "SIII" performance people seem to think they will get. You got a single core processor and 512mb of RAM so it is what it is. Don't get me wrong, a FEW tweaks are helpful but if you overdo it they will all cancel out each other.
Sent from my SGH-T679 using xda app-developers app
Click to expand...
Click to collapse
512 MB of ram you say? Is that what it's supposed to be under the memory spot in the about section in settings?
Sent from my SGH-T679 using xda app-developers app

#1n00b said:
512 MB of ram you say? Is that what it's supposed to be under the memory spot in the about section in settings?
Sent from my SGH-T679 using xda app-developers app
Click to expand...
Click to collapse
No it should say 356. Part of that 512 is not available for use. So in reality we have even less than what's physically there. That is the real problem with this phone. Most of the freezing and lagging is because of the lack of sufficient memory. No amount of memory tweaks help very much either. The OC kernels do help a bit but nothing makes up for not having enough memory. Just like an old PC from 10 years ago that's trying to run Windows 7, think of it that way.

ArtfulDodger said:
No it should say 356. Part of that 512 is not available for use. So in reality we have even less than what's physically there. That is the real problem with this phone. Most of the freezing and lagging is because of the lack of sufficient memory. No amount of memory tweaks help very much either. The OC kernels do help a bit but nothing makes up for not having enough memory. Just like an old PC from 10 years ago that's trying to run Windows 7, think of it that way.
Click to expand...
Click to collapse
Man I was worried for a second.
Sent from my SGH-T679 using xda app-developers app