Hi im thinking of doing a overclock to 1.5 ghz is it safe for the processor and the battery life witch already sucks. has anyone on here whos done it using setcpu and the rom had any problems after doing so ? and how much faster is 1.5 ghz compared to 1.2 ghz on the galaxy s2 ? and is it safe in general for the hardware of the phone will the phone get hotter or have over heating problems ? also what should my settings be for minimum maximum and the voltages ?
Of course oc is going to make ur battery life a lot worse.
Sent from my Inspire 4G using xda premium
I haven' used Setcpu in a while but I believe it's got a feature to stress test the settings to see if it can handle it. Generally the worst that happens is it'll crash. So long as it doesn't get hot. Heat=death for electronics :')
I'm using tegrak (from market) to overclock to 1.4. I found 1.5 was too unstable. Noticeable difference between 1.2 & 1.4 is almost none. Also voltage increase of 50mv was needed to gain stability. There is not much point of overclocking the device IMO unless like me u just get some satisfaction out of knowing its running faster!
Sparksltd said:
I'm using tegrak (from market) to overclock to 1.4. I found 1.5 was too unstable. Noticeable difference between 1.2 & 1.4 is almost none. Also voltage increase of 50mv was needed to gain stability. There is not much point of overclocking the device IMO unless like me u just get some satisfaction out of knowing its running faster!
Click to expand...
Click to collapse
This is true. I see no difference between 1.2 & 1.5ghz. But that doesn't keep me from running it at 1.5 when I'm playing a game anyway :'D
Sent from my Inspire 4G using xda premium
GMoneyDTP said:
Hi im thinking of doing a overclock to 1.5 ghz is it safe for the processor and the battery life witch already sucks. has anyone on here whos done it using setcpu and the rom had any problems after doing so ? and how much faster is 1.5 ghz compared to 1.2 ghz on the galaxy s2 ? and is it safe in general for the hardware of the phone will the phone get hotter or have over heating problems ? also what should my settings be for minimum maximum and the voltages ?
Click to expand...
Click to collapse
Well it's not safe, that I can say. Nothing is safe when you don't know what other things are changing with the clock rate. But still you could do that. You just have to make sure that the temperature is within limit, it will be definitely high, but make sure it's not that high.
And upping the clock rate to 1.5 GHz won't drain any extra battery than 1.2 GHz.
Regards.
GMoneyDTP said:
Hi im thinking of doing a overclock to 1.5 ghz is it safe for the processor and the battery life witch already sucks. has anyone on here whos done it using setcpu and the rom had any problems after doing so ? and how much faster is 1.5 ghz compared to 1.2 ghz on the galaxy s2 ? and is it safe in general for the hardware of the phone will the phone get hotter or have over heating problems ? also what should my settings be for minimum maximum and the voltages ?
Click to expand...
Click to collapse
I tried the overclocking and each phone is different, increase then test, there are many benchmarking apps out there, I was stable at 1.4, my kernel does not include 1.5 GHz, and irregardless of what people may say, running the phone at higher clock speeds consistently DOES drain the battery more than 1.2GHz, unless you have found a way to manipulate the laws of physics. I don't see how one can fathom the thought that you can get a faster CPU at absolutely no cost.
Many people think when the CPU running faster means you complete tasks faster, hence 1.5 would complete a task faster than 1.2 and hence save battery, but I believe we use this phone as more than a calculator, so when you are performing a task that causes the CPU to run at 1.5 for a period of time e.g playing a game, browsing or whatever you do with your phone, then the CPU has to draw more power to keep the frequency at 1.5 as compared to the CPU running at 1.2
All in all, I dialed back because I was doing fine without it and I did not need it, other people may need it, and I wasn't gaining anything. to answer your question directly, it all depends on what you do with your phone, I'm not a huge gamer or anything and the 1.2GHz has proven to be more than enough for me to have a very pleasant experience using this device
Using it occasionally is safe, unless you stress your mobile at 1.5+ ghz for more than an hour or so.
The cpu governor will only make use of higher frequencies when there is a need, so mostly you will see your mobile idling at 200 mhz.
HTCinspiration said:
Of course oc is going to make ur battery life a lot worse.
Click to expand...
Click to collapse
Not necessarily a lot worse. It depends on how you are using your phone, and what is causing it to run fast. If overclocking allows the phone to spend longer in deep sleep state, your battery will win. If its games, and by OC you achieve a higher frame rate, then the phone is doing more work, and this will be reflected in the battery life going down.
Most of the time, screen on is going to dominate. If you are a low screen-on user, I guess there is more chance for a higher max_cpu to be beneficial (it won't be used unless there is a task that requires it). More accurately, it helps some of the time, so the net result is not always obvious to predict.
How about undervolting?
Any danger in doing so- other than crashing when voltage is too low?
Before over clocking it might be worth downloading cpuspy from the market and seeing what time you spend in each speed.
Chances are most of the time you won't even hit 1.2 so might not have anything to gain by over clocking.
Then again I dont know how you use your device but I do know I was surprised when I found out.
Sent from my GT-I9100 using XDA App
I've played with the SuperSonic version of this phone (which has a 1.5Ghz core) and the results from photos/video are much better, I'm pretty sure the cameras are exactly the same but the SuperSonic benefits from the higher CPU.
I think it would be cool if somebody could write an app that states
IF user opens Camera application voltage + clock speed is increased to allow stable 1.5GHz speed
WHEN user closes camera application overclock turns off.
Thoughts?
You have a smartphone! Not a pc!
I didnt see any advantage if you oc to 1.5
Btw In normal use your phone keep in 200mhz not 1.2.
I think i only hit 1.2 two or 3 times for 3 seconds in a full battery cycle.
Sent from my GT-I9100 using xda premium
oc dosnt make sense ,there no noticeable diff,an yes very few times i hit 1.2 mostly its on 800 or few times on 1000....using a device with 1.5ghz proccesor an oc a device to 1.5 ghz which has 1ghz processor is very diff in nature....
Fair enough.
Well I wouldn't mind an app that mazes out at 1.2GHz with camera app open, as it can lag when shooting HD.
oc upto 1.4 ghz is fair enough, anything more than that is a waste. though you cant find any visible difference between 1,1.2 and 1.4 it does helps in maintaining smoothness on some situations.
yes very true i had oc my sgs1 at 1.2ghz and sgs2 at 1.4ghz but frankly i dint see the diff so after few days use i turnd to orignal settings....but many use oc an phone works fine,but depends on personel use......an yes as bala_gammer says its fine on 1.4ghz......cheeeerz
is overclocking galaxy s2 to 1.3gh safe
HEY pople and friends i want to overclock my samsung galaxy s2 to 1.3gh soo is it safe ??????.
imad.d said:
HEY pople and friends i want to overclock my samsung galaxy s2 to 1.3gh soo is it safe ??????.
Click to expand...
Click to collapse
hey mate
u really think 100mhz would make that much change? its safe btw... seen people running s2 at clock speed higher than 1.3ghz
Plz help @..Longtime hd gaming(gta sa,nfs mw , asphalt,modern combat...) using 1.4 ghz is safe?
I also noted that gameloft games are makes more over heat for phone..
Which is the best and safe voltage level for 1.4 ghz 1250mv or 1275mv or 1300mv
Iam using siyah v6.0b5 kernel with wizzedkat 3.1..
Sent from my GT-I9100 using XDA Premium 4 mobile app
Well u do not see much difference on daily usage due to oc..but it does give a smoother experience while playing high quality games
Sent from my GT-I9100 using Tapatalk
I figured I would run quadrant stand and see what I got and I got some interesting results. I got a score of 1830 which is a little bit less then what people have gotten on my phone the Samsung Galaxy Prevail which is a lowend single core 800 MHz Qualcomm MSM7627-3 plus the phone only has 384megs of ram. On top of that quarant says the device only has 1 core so Im not sure whats going on with that
I really don't care for quadrant -- at all. It provides far too divergent results on different devices of the same model/SW, it often does not use all available resources to test a device, and I find its results far too inconsistent to be worth regarding as anything other than a random grouping of numbers.
Im just wondering what other apps see the cpu as 1 core and not the dualcore it really is like quadrant does.
This section is not for discussions like this. This is more suited to the general section.
Moved to General
Montisaquadeis said:
Im just wondering what other apps see the cpu as 1 core and not the dualcore it really is like quadrant does.
Click to expand...
Click to collapse
The Nook Tablet usually turns off the second core in order to conserve battery power. I don't know what the conditions are wherein it will spin up the other core.
hmm could be why some apps are a bit slower then others. and why a lowend smartphone is getting better quadrant scores then this dualcore of a monster
Montisaquadeis said:
I figured I would run quadrant stand and see what I got and I got some interesting results. I got a score of 1830 which is a little bit less then what people have gotten on my phone the Samsung Galaxy Prevail which is a lowend single core 800 MHz Qualcomm MSM7627-3 plus the phone only has 384megs of ram. On top of that quarant says the device only has 1 core so Im not sure whats going on with that
Click to expand...
Click to collapse
Mine got 2140. And a phone with a 800 mhz processor cant reach that unless it is rooted and overclocked. Only on CPU i got around 1500 points. So that means with custom rom the graphics can be improved and with overclock i think it will be able to reach at least around 3500. And quadrant its not so accurate either.
If you check this thread you will see a couple of people that are running CM7 have gotten 1900-2100 scores
http://androidforums.com/galaxy-prevail-all-things-root/479489-quadrant-score.html
Hello,
im a new Galaxy S2 (international version) owner.
I remember quadrant reporting 1 cpu on the stock os (2.3.4) aswell, but didnt try other apps.
Rooted, then updated my device to Android 4.0.3 (ICS ressurection remix v9.6) using the safest methods avalible today (runme.bat + flash cwm via odin + flash rom via cwm)
The ROM includes this kernel: i9100XXLPH
The following applications are now reporting 1 cpu core:
Quadrant
CpuSpy
Elixir 2
Linpack
native "top" utilily, via a terminal emulator
native /proc/cpuinfo
Linpack scores 70+ MFLOPS at 1.2Ghz (multi-thread)
I can hit 4000k points on Quadrant Standard Edition aswell
Is this normal? Im confused. It's the dualcore exynos, but everything reports 1 core.
Anyone? Great forums. Cant post to real threads and here this thread just dies?
Your life must be so awesome when the biggest problem is that your benchmark programs are reporting one core.
Mine correctly reports two, when it's using enough CPU to turn it on.
Not that I ever noticed a performance difference even on GB if I forced one core offline.
Please don't make the exact same thread in two sections of the board. That won't endear you to anyone on here.
The 2nd core pops online when needed, so won't show all the time, that's normal.
Sent from my GT-I9100 using Tapatalk
Hello everybody,
My friend and me(we have s-on and ics stock last version) when we tried quandrant we usually get 2700 scores.But he told me yesterday he tried different way that he added one line in "buid.prop" and he got 3600 marks quandrant scores
When he opened quandtrant benchmark system information he saw just 1 core is writing.He thought we should see 2 cores because we have 2 cores and ics (even leak)version.
So he wants to added in build.prop ''persist.sys.ui.hw=1'' and he wiped dalvic cache and cache and fix permission then restart system.He checked quandtrant standarts and he saw that 2 cores were writing there and immediately he got run full benchmark and got 3600-3500 marks.So we think that the reason of 2 cores.
But when he checked after 20 min later it was writing again 1 core in quantrand system information.It s chancing by itself.we can get more performans? is it possible to keep it?
I remember some guys talking about cores and they told just check antutu you gonna see 2 cores.Second core works when need it.But my friend added ''persist.sys.ui.hw=1" this line in build.prop and maybe he made 2 cores active So what do you think about it?Maybe somebody try this way and share results with us?
"sorry about my english.it is not perfect "
musti95 said:
Hello everybody,
My friend and me(we have s-on and ics stock last version) when we tried quandrant we usually get 2700 scores.But he told me yesterday he tried different way that he added one line in "buid.prop" and he got 3600 marks quandrant scores
When he opened quandtrant benchmark system information he saw just 1 core is writing.He thought we should see 2 cores because we have 2 cores and ics (even leak)version.
So he wants to added in build.prop ''persist.sys.ui.hw=1'' and he wiped dalvic cache and cache and fix permission then restart system.He checked quandtrant standarts and he saw that 2 cores were writing there and immediately he got run full benchmark and got 3600-3500 marks.So we think that the reason of 2 cores.
But when he checked after 20 min later it was writing again 1 core in quantrand system information.It s chancing by itself.we can get more performans? is it possible to keep it?
I remember some guys talking about cores and they told just check antutu you gonna see 2 cores.Second core works when need it.But my friend added ''persist.sys.ui.hw=1" this line in build.prop and maybe he made 2 cores active So what do you think about it?Maybe somebody try this way and share results with us?
"sorry about my english.it is not perfect "
Click to expand...
Click to collapse
The first few pages of this threadhttp://forum.xda-developers.com/showthread.php?t=1616669 talked extensively about dual-core and quadrant scores, and I would say by page 10 it was agreed upon that quadrant really just sucks. Go check it out, maybe that will shine some light for you..
I remember reading about how ICS improved dual core support, and I could have sworn it was on developer.android.com but I can't find the article that explains it now(and too many results to sift thru) - but basically the tweak "persist.sys.ui.hw=1" would have been found in gingerbread roms, because it DID increase performance, however with the way ICS handles the threading of dual-core+ processors that shouldn't do anything.
And before someone comes in here and says this in a not-as-nice way, be careful using quadrant benchmark as a tool- if ONE little detail is different, you will get TOTALLY different results! A different launcher, a different setting, or even having different background processes or having recently run a different app will change the score TOTALLY. Run it 10 times, chances are it will put out wildly different numbers - now go run 10 apps, then go back into quadrant and run it 10 more time - again, wildly different numbers, no consistency.
Besides, all that REALLY matters is your experience - if your score goes up a thousand(consistently somehow) but it doesn't work noticeably smoother, faster, or better- then what does your score matter? You could add 2 more cores, another gig of ram, better gpu - but if you can't NOTICE a difference then what would the point be?
It is easy to find why people added 'persist.sys.ui.hw=1' to gingerbread, but I can't find ANY documentation/posts that give a real reason to add it in an ICS build.. hope this answers your question.
And if anyone can find the link to googles explanation on how ice cream sandwich handles multi-core cpus compared to gingerbread can you PM it to me?
Thanks my friend your answer.Just ı got that if use quandrant benchmark test when ı check system information if ı see 2 core ı get 3500 marks but if ı see just 1 core system informatin ı can get 2700 marks.
So it depents on cores ı see.Its fault of benchmark app or our phone works one core just when need it use second core?
Device - Samsung Galaxy S4
ROM - Resurrection Remix 6.0.1_r72 (https://forum.xda-developers.com/galaxy-s4/i9505-develop/rom-resurrection-lollipop-v5-5-5-t3195202)
Kernel - Alucard kernel repacked from https://forum.xda-developers.com/ga...p/jdcteam-optimized-cyanogenmod-14-0-t3479888
Rooted
There are no problems with software but I've put it here anyway as the instructions suggest.
My problem is the following - I've been playing with undervolting to make my battery last longer and in order to test the stability I've ran StabilityTest.
The voltages seems to be stable but I've noticed a big problem with throttling once I get into 1V domain.
The cores behave totally separately as if on a different board instead inside the same processor.
The first thing is that on idle 3 out of 4 cores will report 32-35C but the last core shows 0.
The second thing happens under a full load - The first core gets super hot very fast (up to 70 when it begins to throttle) the cores 2&3 are around 40-50C and the last core is still stubbornly locked on 0 even though it's under a full load!
It seems obvious to me that the temperature is not reported correctly but I don't know what to do about it.
Is this about thermal sensor being moved out of place or broken? Or something else?
Any advice is welcome.