Originally from this thread.
3 voltage tables exist in our kernel. Which table is used is based on QFuses that are blown. Based on the code, there are 4 "classes" of CPUs: Slow, Nominal, Fast, and Faster. The decision on which table is used is based on results from reading QFuses. I am going to make the ASSUMPTION that some testing happens at manufacturing/testing stage and fuses are blown based on the results.
The tables dictate a 50mv difference from slow to nominal and another 50mv from nominal to fast.
Example: @ 384mhz, a "slow" cpu is @ 950mv, a "nominal" cpu is @ 900mv, and a "fast" cpu is @ 850mv.
You can find how your CPU is binned by booting into (non-stock) recovery and running:
adb shell dmesg | grep PVS
Click to expand...
Click to collapse
Remember to run adb shell, hit enter, and then run the rest. It's not a huge deal, but it's still kind of interesting. I have a nominal phone, and I can OC to 2 GHz with just a little extra voltage.
nominal oh how i wish it was faster.. but i guess its better than slow right
jjblaster3 said:
nominal oh how i wish it was faster.. but i guess its better than slow right
Click to expand...
Click to collapse
Nominal buddies!
I did this on my nexus 4 a while ago, mine is fast. Chongos is faster, and xs has fast too. My hox is liquid plastic now so I can't do this
Sent from my Nexus 4
RollTribe said:
Nominal buddies!
Click to expand...
Click to collapse
hehe yeah funny thing though, switched from bulletproof to elementalx 6.5 today, and found out my phone was able to handle 800mv settings (which applies to 192 and 384 mhz) so im wondering how the slow nominal and fast concept fits in here lol.
i should mention that previous version (5.4) was unstable at 850mv, but had only tried it once though before going to bulletproof..
mirGantrophy said:
My hox is liquid plastic now so I can't do this
Click to expand...
Click to collapse
sorry for your loss..
i hope you dont mean "liquid plastic" in the literal sense, if so how'd it happen
jjblaster3 said:
sorry for your loss..
i hope you dont mean "liquid plastic" in the literal sense, if so how'd it happen
Click to expand...
Click to collapse
He set it on fire out of hate for HTC, and also because it wasn't usable in any state what so ever than a paperweight.
Sent from my HTC One X using xda premium
jjblaster3 said:
hehe yeah funny thing though, switched from bulletproof to elementalx 6.5 today, and found out my phone was able to handle 800mv settings (which applies to 192 and 384 mhz) so im wondering how the slow nominal and fast concept fits in here lol.
i should mention that previous version (5.4) was unstable at 850mv, but had only tried it once though before going to bulletproof..
sorry for your loss..
i hope you dont mean "liquid plastic" in the literal sense, if so how'd it happen
Click to expand...
Click to collapse
Myrder said:
He set it on fire out of hate for HTC, and also because it wasn't usable in any state what so ever than a paperweight.
Sent from my HTC One X using xda premium
Click to expand...
Click to collapse
Actually it caught on fire in my hands so I ran it over with my truck
Sent from my Nexus 4
My phone is Fast. Does that mean I can undervolt it??
r3xx3r said:
My phone is Fast. Does that mean I can undervolt it??
Click to expand...
Click to collapse
It means that it should require less voltage than most phones to be stable, so yes, you should be able to undervolt further.
grep PVS returns 0 lines for me.
I outputted the entire dmesg into a file...
http://pastebin.com/LfutkCkB
mirGantrophy said:
Actually it caught on fire in my hands so I ran it over with my truck
Sent from my Nexus 4
Click to expand...
Click to collapse
For real ?
Sent from Xparent Red using my Venomized Evita
gregneal said:
grep PVS returns 0 lines for me.
I outputted the entire dmesg into a file...
http://pastebin.com/LfutkCkB
Click to expand...
Click to collapse
Did you run the first part of the command?
RollTribe said:
Did you run the first part of the command?
Click to expand...
Click to collapse
...did you read the whole of my post?
of course i did haha, i know a little about unix cmds. dmesg is piped to grep, which will only return lines containing the word PVS.
gregneal said:
...did you read the whole of my post?
of course i did haha, i know a little about unix cmds. dmesg is piped to grep, which will only return lines containing the word PVS.
Click to expand...
Click to collapse
Lol sorry I just glanced over it. I was just assuming you were a noob. I don't know what to tell you other than just reboot into recovery and try again. If you really must know then you probably should head over to the original thread
I'm on viperxl latest and it does same thing after cmd it just goes to line below which is blank. I tried terminal in android as well as in recovery. I ran adb shell then did next cmd. My phone has its own mind and it didn't like me peeking under dress...
Sent from my HTC One X using xda premium
RollTribe said:
Lol sorry I just glanced over it. I was just assuming you were a noob. I don't know what to tell you other than just reboot into recovery and try again. If you really must know then you probably should head over to the original thread
Click to expand...
Click to collapse
No worries. i just booted into recovery again and used the TWRP terminal rather than going over ADB, and it worked. Nominal for me as well
Just ran it from TWRP terminal command and I have a fast Binned CPU
Faster ftw
I figured uv'ing to 800 was pretty normal. Learning something new every day.
Mines on fast. Curiosity got to me
Mines slow
Sent from Xparent Red using my Venomized Evita
Related
Take a look at this link
The GPU has been overclocked to 200 MHZ result from stock 25 fps to 36.5...
The desire(+-27FPS neocore) has the AMD Z430 GPU wich is also on the acer Liquid running at much higher 34,6 FPS @ the same resolution 800x480 but UC cpu768 MHZ although it still preforms much higher on neocore !
I'm sure that some of you developers with the knowledge are prepared figuring this out to get the desire GPU preforming better !..
(tnx to beidl) for those who're trying to find the changes: http://git.linuxtogo.org/?p=groups/m...8a6a0dfe5ba912
The Snapdragon Processor holds the CPU and GPU in one chip... i dont know if youre able to do that...
EDIT: maybe changing some clocking and voltage values in kernel???
More performance can somehow be squeezed out of it tho. Its just not good enough with the power it holds
-------------------------------------
Sent via the XDA Tapatalk App
I Agree there is more..
and it will drain the battery some more...
pina_ said:
and it will drain the battery some more...
Click to expand...
Click to collapse
I don't think people that wants to get the best benchmarks.... care? lol goto maximum overclocking.... possilbey higher than 1.27GHz..... after you benchmarked and got your name in the list.... go back to UVOC lol
atomfix said:
I don't think people that wants to get the best benchmarks.... care? lol goto maximum overclocking.... possilbey higher than 1.27GHz..... after you benchmarked and got your name in the list.... go back to UVOC lol
Click to expand...
Click to collapse
Yeah thats why im no 1
Come get me mate lollol
JD
JupiterDroid said:
Yeah thats why im no 1
Come get me mate lollol
JD
Click to expand...
Click to collapse
lol, well if Richard would like to make a temporary kernel that would restore the volt and try 1.4GHz maybe more. HTC Desire could be getting the No.1 spots in benchmarking
for those who're trying to find the changes: http://git.linuxtogo.org/?p=groups/...;hpb=110e73c591db3fd23d8558659f8a6a0dfe5ba912
looks like only some parameters and 2 functions (pc_clk_enable and pc_clk_disable) were changed.
I don't have much experience with kernel developing, I'm just a stupid java dev. lol
But I'd like to see this added to the 2.6.32.9 kernels, and since we would be able to
change the clock frequency somewhere in /sys/module/clock/, something like a "SetGPU" would be a nice project to do.
beidl said:
for those who're trying to find the changes: http://git.linuxtogo.org/?p=groups/...;hpb=110e73c591db3fd23d8558659f8a6a0dfe5ba912
looks like only some parameters and 2 functions (pc_clk_enable and pc_clk_disable) were changed.
I don't have much experience with kernel developing, I'm just a stupid java dev. lol
But I'd like to see this added to the 2.6.32.9 kernels, and since we would be able to change the clock frequency in /sys/module/clock_7x00/parameters/gpu, something like a "SetGPU" would be a nice project to do.
Click to expand...
Click to collapse
That would be sweet
Sounds good to me Im up for throwing my Desire in the firing line
JD
JupiterDroid said:
Sounds good to me Im up for throwing my Desire in the firing line
JD
Click to expand...
Click to collapse
same haha
i'd totally do it too....
*Hopes that richardtrip will take a look at this thread* lol
i wont be happy till were above the n1 in linpack
and ...
till we have 1800 in quadrant (n1 best so far is 1750)
Stay on topic please!!
i wont be happy till were above the n1 in linpack
and ...
till we have 1800 in quadrant (n1 best so far is 1750)
Click to expand...
Click to collapse
True... we need to get ahead of N1 lol I'm ashamed of the performance it already shows lol
-------------------------------------
Sent via the XDA Tapatalk App
How many batteries do you wanna carry?
eezray said:
How many batteries do you wanna carry?
Click to expand...
Click to collapse
Solar Charger
Fstluxe,
First of all Thank you!! this is the Time and way to Inspire the devs that have the knowledge ,letting them bring this amazing Patch to the desire and maybe all the android Phones ! Great thank you for awaking the community..
So developers Please Come on Just try this out please!! i will putt my money on THIS project!!!!!
I have found this hack in the Thunderbolt android development thread in some ROM, so I decided to shared with you guys
- Download the attached file. Unzip the file.
- Push the unzipped file in xbin
adb push uber_quadrant /system/xbin
adb shell chmod 755 /system/xbin/uber_quadrant
- Install Quadrant (Advanced or Basic)
- Execute the pushed file.
adb shell /system/xbin/uber_quadrant
- Run Quadrant and be amazed by the score.
Mine is attached without any overclocking.
What does this actually do?
Aliq said:
What does this actually do?
Click to expand...
Click to collapse
It hacks the application and gives you amazing scores.
Try it out
theintelligent said:
It hacks the application and gives you amazing scores.
Try it out
Click to expand...
Click to collapse
Yeah but it isn't useful in any other way, because its not the real score..
Sent from my Desire HD using Tapatalk
2271 without OC? That is how it is supposed to perform, a fully stock DHD can do that.
just shows how little value benchmarking has
hoseth0711 said:
Yeah but it isn't useful in any other way, because its not the real score..
Sent from my Desire HD using Tapatalk
Click to expand...
Click to collapse
Yes exactly that is the point, these benchmarks should not measure performance of any ROM rather its usability should be the judge.
This is what is in the file:
#!/system/bin/sh
mount -o rw -t tmpfs tmpfs /data/data/com.aurorasoftworks.quadrant.ui.standard
mount -o rw -t tmpfs tmpfs /data/data/com.aurorasoftworks.quadrant.ui.advanced
Click to expand...
Click to collapse
capychimp said:
just shows how little value benchmarking has
Click to expand...
Click to collapse
why does it shows this? i see absolutely no relation to benchmarking. this is faking results. benchmarking is an objective messure of a speciffic (or a range of specific) performance aspects. in fact benchmarking is the only way to messure improvement.
the problem rarely lies wihtin benchmarking but most time at the user that does not understand the benchmark and false concludes from the results.
but in this case this has absolutely nothing to do with the often heared critics about benchmarks.
edit: the script that gets executed simply tricks the io system benchmark.
jkoljo said:
2271 without OC? That is how it is supposed to perform, a fully stock DHD can do that.
Click to expand...
Click to collapse
LOL true story.
before i flashed mines, i had 2250 score with stock rom.
hmm, threadstarter self pwn with this thread?
theintelligent said:
It hacks the application and gives you amazing scores.
Try it out
Click to expand...
Click to collapse
Overclocked or not, that scores weak my friend, i've had my dhd get over 3000 OC'd.
For whatever reason the advanced quadrant gives you better scores, however neither of them offer an actual reallife score which means your phone still runs the same, kind of pointless, no offence.
I've never really understood the fascination with benchmarks so excuse my ignorance, but is this not like writing 300mph on the speedo in your car and then getting all excited when the needle hits it?
jkoljo said:
2271 without OC? That is how it is supposed to perform, a fully stock DHD can do that.
Click to expand...
Click to collapse
True!
I don't see the point of benchmarking. So, I really don't see the point in faking IT!
Lol, this is still a good thread to laugh
Sent from my HTC Desire HD using Tapatalk
CitizenLee said:
... is this not like writing 300mph on the speedo in your car and then getting all excited when the needle hits it?
Click to expand...
Click to collapse
Exactly! There are some very sad people about.
my DHD hardly reaches 1350 with Quadrant standard (stock rom), i wonder how some ppl are achieving 2500 with stock...
swehbe said:
my DHD hardly reaches 1350 with Quadrant standard (stock rom), i wonder how some ppl are achieving 2500 with stock...
Click to expand...
Click to collapse
2500 is not possible with stock...
anyway, to get 2000 and a bit above you have to switch the cpu governor to performance. to do this you need root and setcpu. with the default ondemand governor this is not possible.
but do not keep performance governor for daily use! it keeps your cpu freq at max all the time.
Lmao! That's hilarious!
Sent from me LeeDrOiD loaded Desire HD using the Android XDA Premium app
swehbe said:
my DHD hardly reaches 1350 with Quadrant standard (stock rom), i wonder how some ppl are achieving 2500 with stock...
Click to expand...
Click to collapse
You should get a new rom!
My DHD just did 2534 using LeeDroid's kernel at 1.2ghz
'Amazing' doesn't really belong in the title of this thread.
Just got my second G2x yesterday got it rooted and bloat ware froozen now I can't remember where I go from 1200 mah to 1500 mah. Can anyone refresh my memory
sgt_addyjp said:
Just got my second G2x yesterday got it rooted and bloat ware froozen now I can't remember where I go from 1200 mah to 1500 mah. Can anyone refresh my memory
Click to expand...
Click to collapse
1500 isn't necessary. In fact, you're risking frying your phone! 1200 is the median for most.
Common Sense is learned..........
do 1500 but when u processor will be overheat, your phone gonna restart.
thats all....I don't remember no one in my life who broke the PHONE because overclocking.
For me 1400 best...because 1500 not stable for "heavy congestion."
avetny said:
do 1500 but when u processor will be overheat, your phone gonna restart.
thats all....I don't remember no one in my life who broke the PHONE because overclocking.
Click to expand...
Click to collapse
It happens often. Why do you think is called "OVERclocking"? OEM's set specific numbers on their kernels for a reason. Besides the numbers themselves and meaningless quadrant scores, you'll probably not notice any difference between 1200 or 1400
Common Sense is learned..........
Chance Ill said:
It happens often. Why do you think is called "OVERclocking"? OEM's set specific numbers on their kernels for a reason. Besides the numbers themselves and meaningless quadrant scores, you'll probably not notice any difference between 1200 or 1400
Common Sense is learned..........
Click to expand...
Click to collapse
oh yea? play flash on 1.2Ghz and 1.4Ghz
avetny said:
oh yea? play flash on 1.2 and 1.4
Click to expand...
Click to collapse
I use flash literally everyday! No problems whatsoever and I'm using stock kernel at 1ghz
Common Sense is learned..........
Chance Ill said:
I use flash literally everyday! No problems whatsoever and I'm using stock kernel at 1ghz
Common Sense is learned..........
Click to expand...
Click to collapse
Can I use your phone for a minute
I will show ALL difference
avetny said:
Can I use your phone for a minute
I will show ALL difference
Click to expand...
Click to collapse
Lol! We can agree to disagree, but 1500 on any kernel is unstable at best!
Common Sense is learned..........
Chance Ill said:
Lol! We can agree to disagree, but 1500 is on any kernel isunstable at best!
Common Sense is learned..........
Click to expand...
Click to collapse
yes, totally agree with you!!!
Chance Ill said:
I use flash literally everyday! No problems whatsoever and I'm using stock kernel at 1ghz
Common Sense is learned..........
Click to expand...
Click to collapse
+1 on stock 1 ghz I can do anything I want without lag lol. I have oced but noticed little to no difference in anything I do. Love this phone. Benchmarks really don't mean much to me when I can't see any difference in my hand.
Sent from my G2x
Chance Ill said:
Lol! We can agree to disagree, but 1500 on any kernel is unstable at best!
Common Sense is learned..........
Click to expand...
Click to collapse
Good dialogue, but he was asking 'where' to change the settings, not if he should change the settings
alodar1 said:
Good dialogue, but he was asking 'where' to change the settings, not if he should change the settings
Click to expand...
Click to collapse
yes I had my other G2x set to 1500 and the battery life was better for me...I prolly set it to 1400 to see how things go. I had to get a new G2x because say someone set me a text msg saying "hey" i would get 200-400 of them and then one day I restarted it and it go to the point where it said LG with the blue bar under it then it would restart...why I have no clue but I would like to figure out why before I have to send it back because I have it rooted and would like to unroot it before sending it back. I also have pictures on the phone that I don't really want to lose so if anyone could help me with that that would be great also.
nobody is really going to answer his question? If you are trying to change the speed either download a program called pimp-my-cpu or setcpu. Both of those will give you full access to do what ever u please to your phones processors. Hope this helps!
there is a place in the setting that you can do it at i just don't remember where to find it at.
Tobenamed-Kernel_test14.zip
http://androidforums.com/triumph-all-things-root/401243-twisted-triumph-kernel-testing.html
http://www.mediafire.com/?28ula4093x527so
And sorry im new to posting on here
Dude... this kernel is f'ng sexy. Did u just fix all my response issues? So far so good. Appreciate the work you and anyone else did to get this out. I'm using v1.2. 1.1 wouldn't boot but this seems great
MOTWX435KT
Using v1 and it runs just fine. This is a good day.
Hey gary, response issues with the kernel? If uve got any probs let me know cuz I'm the dev
Sent from my PG06100
treyj5 said:
Tobenamed-Kernel_test14.zip
http://androidforums.com/triumph-all-things-root/401243-twisted-triumph-kernel-testing.html
http://www.mediafire.com/?28ula4093x527so
And sorry im new to posting on here
Click to expand...
Click to collapse
this is the kernel i use.. and it works great for me.. 1.8ghz and stable.
gravityone said:
this is the kernel i use.. and it works great for me.. 1.8ghz and stable.
Click to expand...
Click to collapse
yea thats my kernel..lol u said response issues i thought u meant with the kernel
and also treyj5 when i get a version going for the "oddball" devices that cant boot v1 im probably gunna post it here, when that happens if u could have a mod close this thread thatd be cool theres just no point in having two threads about the same thing
Running @ 1400 (Scary) and its super smooth. Thank you! Better battery life too!
Rezanow said:
Running @ 1400 (Scary) and its super smooth. Thank you! Better battery life too!
Click to expand...
Click to collapse
yea scarys my fav gov too, once i get this issue with the oddball devices worked out itl get better battery life cuz im gunna try and implement the UV that doomlords kernel has in v2, but yea id like to get it booting on all devices before i get into that
Response is great. On my evo I used to get certain kernels that would be unresponsive and buggy with touch. feels like the stock triumph kernel doesn't work too well with the rest of the os. This kernel flys
MOTWX435KT
i talked to austrie and when he gets the issues worked out in his rom http://androidforums.com/triumph-all-things-root/403882-princely-rom-beta.html
and i get the odball issues worked out in the kernel, were gunna make special versions and u guys r gunna get a rom on ext4 so good things to come for this device
I am having a problem. My governor keeps returning to ondemand, despite my setting it to scary. I haven't restarted the phone yet. What am I doing wrong?
pwnyourace said:
i talked to austrie and when he gets the issues worked out in his rom http://androidforums.com/triumph-all-things-root/403882-princely-rom-beta.html
and i get the odball issues worked out in the kernel, were gunna make special versions and u guys r gunna get a rom on ext4 so good things to come for this device
Click to expand...
Click to collapse
Best news I've heard in a while. Where's the paypal link? xD
Sent from my rooted Moto Triumph
Rezanow said:
I am having a problem. My governor keeps returning to ondemand, despite my setting it to scary. I haven't restarted the phone yet. What am I doing wrong?
Click to expand...
Click to collapse
ok inside system/ect there is a file called "init.qcom.post_boot.sh" delete it, then reboot and all of your settings will stick.
ziggy46 said:
ok inside system/ect there is a file called "init.qcom.post_boot.sh" delete it, then reboot and all of your settings will stick.
Click to expand...
Click to collapse
Done. All seems well so far.
Fodoom said:
Best news I've heard in a while. Where's the paypal link? xD
Sent from my rooted Moto Triumph
Click to expand...
Click to collapse
I put one up in the Android forums thread, trying to save up and buy this phone for deving cuz they're pretty cheap on Craigslist around me, I sent Austrie that kernel so lets see what he can come up with
Sent from my PG06100
I am running 1900/245 without error, but I am struggling with the different available governors (specifically: ondemand, interactive, smartass, and scary). Can anyone explain the differences in layman's terms?
Sent from my MOTWX435KT using xda premium
So is this kernel fully functioning? No known bugs?
Anyone noticed significant impact on battery life (good or bad)?
maxpower17 said:
So is this kernel fully functioning? No known bugs?
Anyone noticed significant impact on battery life (good or bad)?
Click to expand...
Click to collapse
Depending on which governor you choose. Scary seems to really keep tabs on it well. I'm at 1900/245 and when I run Scary the battery doesn't run down at all. I'm not a fan of the performance though, as it is too slow to react sometimes. Still working on the others, so no solid opinions yet.
Rezanow said:
Depending on which governor you choose. Scary seems to really keep tabs on it well. I'm at 1900/245 and when I run Scary the battery doesn't run down at all. I'm not a fan of the performance though, as it is too slow to react sometimes. Still working on the others, so no solid opinions yet.
Click to expand...
Click to collapse
I don't see why you would underclock your phone that slow... 245 is way too slow for a phone made to run default at 1ghz. You'll have screen wake issues and uber response problems... duh.. I don't see why ppl turn their smartphones into expensive paperweights just so they can extend battery life. I'd rather have a phone that computes efficiently than one that doesn't even want to process. 600 MHz min is just fine.
Sent from my Triumph using Tapatalk
245 would be the minimum clock speed, not the maximum My guess is that when your phone is asleep, it would use as little resources as possible.
Sent from my rooted Moto Triumph
It was suggested to me that in order to prevent excessive battery drain and heat when using CM7 I should disable one of the cores. Has anyone done this? What negative impacts could doing this cause (if any)?
webmaster said:
It was suggested to me that in order to prevent excessive battery drain and heat when using CM7 I should disable one of the cores. Has anyone done this? What negative impacts could doing this cause (if any)?
Click to expand...
Click to collapse
If anything it would increase battery drain due to all the processing being put on one single core, what I would recommend for a decrease in battery drain is different kernels with under volting options. That's what's worked best for me.
Sent from my Sensation using xda premium
Chilled26 said:
If anything it would increase battery drain due to all the processing being put on one single core, what I would recommend for a decrease in battery drain is different kernels with under volting options. That's what's worked best for me.
Sent from my Sensation using xda premium
Click to expand...
Click to collapse
It was my perception that the older single core phones had better battery life. Do you really think it would hurt the battery?
Dual core isn't properly utilized yet so it might not be a bad idea.
Here is an example of my battery drain for today with the bricked kernel v1.4.
Sent from my Sensation using xda premium
Changing kernel and undervolting did the trick for me. Getting 20 hours moderate use.
Sent from my Sensation using xda premium
Chilled26 said:
Here is an example of my battery drain for today with the bricked kernel v1.4.
Sent from my Sensation using xda premium
Click to expand...
Click to collapse
What are your settings?
Most of the time when both cores are on the 2nd isn't doing anything, because the first core isn't even pushed enough to reach max. So, I feel 1 core is sufficient, I'm running 4.1.x w/ 1 core @800 and the battery and speed are fine. Although I see others running speeds of 1.2 and higher, but I barely notice the difference of the freqs. The S3 is so fast that the higher clocks are overkill IMO.
Ace42 said:
Most of the time when both cores are on the 2nd isn't doing anything, because the first core isn't even pushed enough to reach max. So, I feel 1 core is sufficient, I'm running 4.1.x w/ 1 core @800 and the battery and speed are fine. Although I see others running speeds of 1.2 and higher, but I barely notice the difference of the freqs. The S3 is so fast that the higher clocks are overkill IMO.
Click to expand...
Click to collapse
I suspect you see a noticeable difference in battery life then? Also, what method did you use to disable the second core?
webmaster said:
What are your settings?
Click to expand...
Click to collapse
Honestly I don't even touch the settings I just run at default.
Sent from my Sensation using xda premium
Pretty snappy
su
chmod 644 /sys/devices/system/cpu/cpu1/online
echo "0" > /sys/devices/system/cpu/cpu1/online
chmod 444 /sys/devices/system/cpu/cpu1/online
Someone in faux kernel forum was talking about doing this. Gscript app makes it easy to run script. If you dont like it just reboot and 2nd core is back on.
Sent from my Sensation using XDA Premium App
Chilled26 said:
Honestly I don't even touch the settings I just run at default.
Sent from my Sensation using xda premium
Click to expand...
Click to collapse
I tried Faux's and it was ok, I plan on trying Bricked tomorrow. However if disabling a core gives even better results then that might be the way to go.
webmaster said:
I suspect you see a noticeable difference in battery life then? Also, what method did you use to disable the second core?
Click to expand...
Click to collapse
echo 0 > /sys/devices/system/cpu/cpu1/online
I do see a difference in battery life, which is why I prefer this setup, when we have ICS I will use both cores. But I feel cheated on Ginger so I use 1 for now, and whoever said two cores has better batt is wrong...Logically two cores use twice the voltage, that means 1 cores will save a significant amount of power. A weak dual core w/ 1 core on will have bad life, but a fast on is different.
Ace42 said:
echo 0 > /sys/devices/system/cpu/cpu1/online
Click to expand...
Click to collapse
Is that permanent or does it need to be run each time you reboot?
webmaster said:
Is that permanent or does it need to be run each time you reboot?
Click to expand...
Click to collapse
Every boot, you can make a linux script(.sh) if you want, and then use script manager. But, I'm only testing this for comparison, I can't see the speed difference w/ 1 or 2 so I just use 1. 1 @800 core feels how fast my MT4G did when OC'd @1.4Ghz.
Ace42 said:
Every boot, you can make a linux script(.sh) if you want, and then use script manager. But, I'm only testing this for comparison, I can't see the speed difference w/ 1 or 2 so I just use 1. 1 @800 core feels how fast my MT4G did when OC'd @1.4Ghz.
Click to expand...
Click to collapse
How long have you been running like that?
I run Faux kernel, Joe's RCU, undervolted -50mV and underclocked to 1Ghz, I can now easily make it through the day and I'm a moderate to heavy user. I'm pretty sure disabling the second core isn't necessary, and frankly, if you check out how little it's used, it probably wouldn't benefit that much anyhow.
Ace42 said:
echo 0 > /sys/devices/system/cpu/cpu1/online
I do see a difference in battery life, which is why I prefer this setup, when we have ICS I will use both cores. But I feel cheated on Ginger so I use 1 for now, and whoever said two cores has better batt is wrong...Logically two cores use twice the voltage, that means 1 cores will save a significant amount of power. A weak dual core w/ 1 core on will have bad life, but a fast on is different.
Click to expand...
Click to collapse
Completely false. Just because something has voltage across it doesnt mean that it is doing anything. Voltage times current equal power. Using power uses the battery.
You guys need to SERIOUSLY change the way you make decisions. Just because something SEEMS logical to you, doesnt mean it is correct. I see alot of BAD info being pushed around here.
I have CM7 and run Fauxs kernels. I FORCE both CPUs online and run at 1.56Ghz. My standby is around 30mAH...Nuff said.
Something else is messed up. Chase facts, not fiction people.
Another issue? UVing. You guys break so many rules in electrical engineering that it isnt even funny. Ive explained things so many times that I am sick of it. I wont anymore, but will let you know that you have missed some key facts.
Matt
mrg02d said:
Completely false. Just because something has voltage across it doesnt mean that it is doing anything. Voltage times current equal power. Using power uses the battery.
You guys need to SERIOUSLY change the way you make decisions. Just because something SEEMS logical to you, doesnt mean it is correct. I see alot of BAD info being pushed around here.
I have CM7 and run Fauxs kernels. I FORCE both CPUs online and run at 1.56Ghz. My standby is around 30mAH...Nuff said.
Something else is messed up. Chase facts, not fiction people.
Another issue? UVing. You guys break so many rules in electrical engineering that it isnt even funny. Ive explained things so many times that I am sick of it. I wont anymore, but will let you know that you have missed some key facts.
Matt
Click to expand...
Click to collapse
So me having better life w/ 1 core is false? Even though I've compared it w/ 2? And I don't know what your problem is but keep it to yourself, I didn't force you to respond to my post, only sharing my experiences.
webmaster said:
How long have you been running like that?
Click to expand...
Click to collapse
Almost 1 day, battery around 30 still. But I listened to music, internet, sync and played some games so my life was shorten.
Ace42 said:
So me having better life w/ 1 core is false? Even though I've compared it w/ 2? And I don't know what your problem is but keep it to yourself, I didn't force you to respond to my post, only sharing my experiences.
Almost 1 day, battery around 30 still. But I listened to music, internet, sync and played some games so my life was shorten.
Click to expand...
Click to collapse
I didnt say this. It is obvious what I said, so why would you even ask me? Just scroll up pal.
You didnt simply give an experience...You implied that your experience is LOGICAL. It may to you, but it is incorrect. Dont just throw things out there, please. It makes people with backgrounds in power and energy pissed off.
Also, my finding completely debuncts your theory. At night, while I sleep, I see 5mAH draw.
Matt