What's new
DroidForums.net | Android Forum & News

This is a sample guest message. Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!

OverClocking 101

Status
Not open for further replies.
How does one O/C past 1200 Mhz. My droid does not seem to boot w/any kernel past 1200. Only once w/UD2.0 I believe it did. I may be using SetCpu wrong. I tried w/Chev's 1.25Ghz and phone just does not boot up. I am using SDM2.0
 
How does one O/C past 1200 Mhz. My droid does not seem to boot w/any kernel past 1200. Only once w/UD2.0 I believe it did. I may be using SetCpu wrong. I tried w/Chev's 1.25Ghz and phone just does not boot up. I am using SDM2.0

stick with what works. i've tried many 1200mhz kernels and my phone does not like them at all.
 
How does one O/C past 1200 Mhz. My droid does not seem to boot w/any kernel past 1200. Only once w/UD2.0 I believe it did. I may be using SetCpu wrong. I tried w/Chev's 1.25Ghz and phone just does not boot up. I am using SDM2.0

stick with what works. i've tried many 1200mhz kernels and my phone does not like them at all.

I was ready for that response but I was hoping for a miracle:icon_eek:. Thanks for smacking reality into my senses. If its not broken, don't fix it right?:)
 
How does one O/C past 1200 Mhz. My droid does not seem to boot w/any kernel past 1200. Only once w/UD2.0 I believe it did. I may be using SetCpu wrong. I tried w/Chev's 1.25Ghz and phone just does not boot up. I am using SDM2.0

You can attempt to do it with an Increased Voltage kernel. There are some major drawbacks to this though.

1) You could possibly kill the phone.
2) Battery life will be abysmal.
3) Overheating will be the norm.


Now for the reason you can't overclock at that speed. There is at least one group of transistors in your CPU that can't handle the clock speed with the voltages you have been using. To overcome this, you keep cranking up the voltage till you find the proper forward biasing that is needed to maintain proper operation. As you can imagine that becomes very risky. The heat generated could cause the materials in the transistor to structurally fail due to the rapid heating and cooling cycles being introduced. There is also the possibility you could burn out one of the support diodes or resistors to the transistors. I would guess you have better than a 30% chance of causing permanent failure of the device within 30 seconds of booting. If the phone didn't shut down, your chances of failure probably increase 10% every minute after that. The good news to all of this is, there are no kernel makers willing to push the top end voltage beyond a certain point. Which means you should only run into this scenario if you start building your own custom kernels with different voltage tables.
 
You're going to be surprised I think. Can't wait for your analysis!

Sent from my Droid using DroidForums App


First week is done and to be blunt, I am not impressed with conservative. I am getting almost exactly the same battery usage as OnDemand provides me. Of course I am not the average user either. I have already done everything I can to cut battery usage down. Now I have been doing some data sampling of the "time in state" file and I have noticed something interesting concerning the differences between OnDemand and conservative. Which I believe would account for why the average user would see a difference in battery life. Starting tomorrow, I am going to alternate between setups and capture the files to show the difference.

The research continues......
 
You're going to be surprised I think. Can't wait for your analysis!

Sent from my Droid using DroidForums App


First week is done and to be blunt, I am not impressed with conservative. I am getting almost exactly the same battery usage as OnDemand provides me. Of course I am not the average user either. I have already done everything I can to cut battery usage down. Now I have been doing some data sampling of the "time in state" file and I have noticed something interesting concerning the differences between OnDemand and conservative. Which I believe would account for why the average user would see a difference in battery life. Starting tomorrow, I am going to alternate between setups and capture the files to show the difference.

The research continues......

Interesting. Well I wonder if just simply my specific cpu likes conservative better for some reason. Cause once I made the switch life has been great...

Anyway, yeah definitely keep up the research! I'm excited to hear the results

Sent from my Droid using DroidForums App
 
I am interested in doing this how can I start it off I read it but dont understand how to get the program onto my phone.
 
Ok well is there one for it because I am really looking into messing with my droid X to make it perform better i am not happy with the speed at the moment. It has alot of things I like about it but lacks a little bit when it comes to speed.
 
You're going to be surprised I think. Can't wait for your analysis!

Sent from my Droid using DroidForums App


First week is done and to be blunt, I am not impressed with conservative. I am getting almost exactly the same battery usage as OnDemand provides me.

So either there is a difference in performance that you're keeping mum about, or the implication is that you're not very impressed with OnDemand, either, since they're about the same!

Seriously, I'm dying to see what your full report will be. It seems to me that Convservative's Advanced settings would make a better label for it ControlFreak, as it appears you get to tell the kernel at what CPU utilization it should ramp up, and at what it can ramp down--and by what percentage.

Assuming the labels for those values are correct, and actually work, that is. If so, seems like one would want to spend a bit of time dialing it in:

Ramping up at, say, 20% and down at 5% by 25% of the max frequency would seem to be a very high performance setting,
Whereas ramping up at 60% and down at 25% by 5% of the max frequency would actually be truly conservative.

Go Skull!
 
You're going to be surprised I think. Can't wait for your analysis!

Sent from my Droid using DroidForums App


First week is done and to be blunt, I am not impressed with conservative. I am getting almost exactly the same battery usage as OnDemand provides me.

So either there is a difference in performance that you're keeping mum about, or the implication is that you're not very impressed with OnDemand, either, since they're about the same!

Seriously, I'm dying to see what your full report will be. It seems to me that Convservative's Advanced settings would make a better label for it ControlFreak, as it appears you get to tell the kernel at what CPU utilization it should ramp up, and at what it can ramp down--and by what percentage.

Assuming the labels for those values are correct, and actually work, that is. If so, seems like one would want to spend a bit of time dialing it in:

Ramping up at, say, 20% and down at 5% by 25% of the max frequency would seem to be a very high performance setting,
Whereas ramping up at 60% and down at 25% by 5% of the max frequency would actually be truly conservative.

Go Skull!

Actually they perform about the same except when I start up a game that needs to ramp up the CPU like Angry Birds and then I see a tad bit of slow down usually only at the startup.

I believe I am unimpressed because I have already tweaked the hell out of my phone. About the only things I have running in background is WeatherBug (every hour), Email (every 15 mins for one account) and SetCPU (in passive mode). I use autostarts to kill off any task that would auto launch on startup or based on events unless I really need the program to know what the phone is doing.

My big battery draws are screen, data (web and rss) and gaming. And those can't be compensated for by any governor configuration.

I am actually very frustrated with myself because I don't believe I can now properly test "real world" savings for the "average user" under this condition.
 
Ok well is there one for it because I am really looking into messing with my droid X to make it perform better i am not happy with the speed at the moment. It has alot of things I like about it but lacks a little bit when it comes to speed.


Not to get off topic but what ROM are you running? If it is the stock Moto/VZW crap that is the major problem, and not the speed of the chip. I don't run there garbage on my device anymore and was overclocked for awhile. I have now taken it back to stock speed on the custom ROM and am more then happy with it.
 
I believe I am unimpressed because I have already tweaked the hell out of my phone...I am actually very frustrated with myself because I don't believe I can now properly test "real world" savings for the "average user" under this condition.
Yes, but the average user reading this thread is aspiring to be more like you (er, as far as battery usage, anyway). :)

When I had the CPU ramp up at a lower percentage threshold, starting games (Homerun Battle, in my case) was zippy; today I raised the threshold a bit, and I did notice a bit of jerkiness, but once in-game, it was fine.

Still looking forward to your findings.

Thanks--

LDog
 
Status
Not open for further replies.
Back
Top