Database Users

Overclocking N1

You can overclock n1 only to 1.190ghz, while desire hd 1.9ghz and the htc desire Z (G2) 2.0ghz. Does N1 has to old cpu?
-------------------------------------
Sent via the XDA Tapatalk App with my Sexy Nexy

Yes. 1st Gen snapdragon
Sent from my Nexus One using XDA App

if you want to OC you N1 go and OC you Desktop is the best choice

Why would you wanna over clock your phone? I have my N1 clocked @ 691 and works really fast with the MIUI rom and battery performance is better than stock. I'm not a fan of custom rom & rooting but I been pretty pleased so far. overclocking the nexus one will drain your battery like crazy plus the 1st Gen of snapdragons weren't as good with graphics as the A4chips and humming birds.

i have mine underclocked too and it works fine. try going a step further and underclocking it to like 422 when it's sleeping/standby. it'll help your battery

josemedina1983 said:
Why would you wanna over clock your phone? I have my N1 clocked @ 691 and works really fast with the MIUI rom and battery performance is better than stock. I'm not a fan of custom rom & rooting but I been pretty pleased so far. overclocking the nexus one will drain your battery like crazy plus the 1st Gen of snapdragons weren't as good with graphics as the A4chips and humming birds.
Click to expand...
Click to collapse
The connection between clockspeed and power consumption is not as strong as you think. But without a doubt it has an influence. Much more important is the voltage. If you "undervolt" the Nexus One CPU you can even get better battery live with higher clockspeed.
And if you use a tool to change the clockspeed depending on the situation (display on/off, battery % left, workload) and undervolt the cpu you can safe A LOT of juice.
With Wildmonks kernel, MIUI and SetCPU I get a much better lifetime than ever before even though my Nexus runs at 1152MHz.

Actually, the frequency makes a BIG difference in power consumption. Think of it this way - each clock causes changes propagating in transistors, which are the actual power draw. More clocks = more changes = more power drawn. As easy as that.
So, having 10% higher frequency and 10% lower voltage compensates each other.
Nexus has examples that overclock to 1.5GHz when overvolted, like Desire Z and Desire HD (both of those have to be overvolted to go up stable from 1.2GHz). Most of Nexus Ones fail when overclocking and don't reach higher than 1.2GHz, but it might be not because of the CPU, but because of other devices on system board.

Generally, it is only when you change the voltage (which is required to stabilize the higher frequency) that you see noticeable differences in battery life.
Jack_R1 said:
Actually, the frequency makes a BIG difference in power consumption. Think of it this way - each clock causes changes propagating in transistors, which are the actual power draw. More clocks = more changes = more power drawn. As easy as that.
So, having 10% higher frequency and 10% lower voltage compensates each other.
Nexus has examples that overclock to 1.5GHz when overvolted, like Desire Z and Desire HD (both of those have to be overvolted to go up stable from 1.2GHz). Most of Nexus Ones fail when overclocking and don't reach higher than 1.2GHz, but it might be not because of the CPU, but because of other devices on system board.
Click to expand...
Click to collapse

willverduzco said:
Generally, it is only when you change the voltage (which is required to stabilize the higher frequency) that you see noticeable differences in battery life.
Click to expand...
Click to collapse
Ok, some additions required.
Leakage is also dependent on power, and the dependency graph isn't linear - and starts breaking upwards at some point, usually being a tad above the max designed voltage.
Going down in voltage makes leakage change approximately linear, and doesn't affect nearly as much as going up.
Overclocking will draw power just as I noted above - exactly with the same percentage difference - only when the clock is reaching the overclocked area, which happens only when you're playing games or doing CPU-intensive tasks.
Undervolting will affect leakage, which happens 100% of the time.
So yes, when running in dynamically scaled environment, undervolting has more effect than overclocking. On desktop PC, running the same clock frequency constantly, the effect is the same.

Very True. And I wasn't saying that overclocking, while at the same voltage, didn't draw ANY more power... I just am trying to say that (for example in this graph) overclocking only has a small effect on power draw until you actually change the voltage. In that same example, going from 3.4 to 3.8 GHz only adds about 6% current draw while at the same vCore, while going up a similar amount in clock speed.
I'd even wager to say that if you're slightly under-volted and as heavily overclocked as you can go at that given voltage, you'll save some trivial amount of power versus stock because of the fact that voltage affects power draw significantly more than clock speed. I would also wager that if you are at an overclocked speed and are at stock voltage, the amount of current and power draw will be almost indistinguishable to the end user, since things like display will almost always use much more power if the display is on for any appreciable amount of time.
Jack_R1 said:
Ok, some additions required.
Leakage is also dependent on power, and the dependency graph isn't linear - and starts breaking upwards at some point, usually being a tad above the max designed voltage.
Going down in voltage makes leakage change approximately linear, and doesn't affect nearly as much as going up.
Overclocking will draw power just as I noted above - exactly with the same percentage difference - only when the clock is reaching the overclocked area, which happens only when you're playing games or doing CPU-intensive tasks.
Undervolting will affect leakage, which happens 100% of the time.
So yes, when running in dynamically scaled environment, undervolting has more effect than overclocking. On desktop PC, running the same clock frequency constantly, the effect is the same.
Click to expand...
Click to collapse

Jack_R1 said:
Actually, the frequency makes a BIG difference in power consumption. Think of it this way - each clock causes changes propagating in transistors, which are the actual power draw. More clocks = more changes = more power drawn. As easy as that.
So, having 10% higher frequency and 10% lower voltage compensates each other
Click to expand...
Click to collapse
I wouldn't call 10% more peak power consumption big if you take in account that the cpu is only running at the max clock speed a very small amount of time. 90% of the time the device is sleeping anyway and even if it's not you barely need the max clock speed. But if you do you will recognize the difference.
On the other side the reduced voltaged can safe you power all the time.
willverduzco said:
I'd even wager to say that if you're slightly under-volted and as heavily overclocked as you can go at that given voltage, you'll save some trivial amount of power versus stock because of the fact that voltage affects power draw significantly more than clock speed. I would also wager that if you are at an overclocked speed and are at stock voltage, the amount of current and power draw will be almost indistinguishable to the end user, since things like display will almost always use much more power if the display is on for any appreciable amount of time.
Click to expand...
Click to collapse
That's exactly what I experienced.

Pommes_Schranke said:
I wouldn't call 10% more peak power consumption big if you take in account that the cpu is only running at the max clock speed a very small amount of time. 90% of the time the device is sleeping anyway and even if it's not you barely need the max clock speed. But if you do you will recognize the difference.
On the other side the reduced voltaged can safe you power all the time.
Click to expand...
Click to collapse
Yes, you're right, and that's why I corrected myself in my second post. I totally forgot about the frequency scaling.

Off topic, but this is why I love XDA. Rational debate over a subject by intelligent people, where there usually isn't flaming. Thanks added to the two of your posts.

[Q] CPU Explanation

Hello, can someone please explain how my battery life improves when I set minimum cpu clock from 200 to 500?
I usually get 3 hours of screen time, with 500 as minimum I get nearly 4 hours. This difference is significant enough to make me wonder and ponder about how this happens.
It's counter logical, something with a higher minimum should drain faster.. right?

Does both cpu states are working at the same voltage? If yes (or the difference is very small) then the battery saving are from the increased execution speed .

In my own opinion, i dont think it affect much, while more than 70%+ battery drain by screen, unless u use cpu 100% all the time, otherwise i wouldnt too much concern abt that
Sent from my GT-N7000 using XDA Premium HD app

Chip efficiency changes with frequency. We tested this on my other phone (an Atrix- nice phone even now!)
Sometimes the CPU is just more power efficient doing tasks at higher frequencies. In a sense, the processor works faster, but for less time- so although it is running faster and requires more battery power, it completes the task much earlier and uses less power in total for that process.
My note, when running at 1.4ghz uses battery so much faster than at 1ghz, but the battery saving when dropping down from 1ghz is minimal, if existent at all.

Welcome to the not-always-intuitive world of modern CPU power usage.
The old mantra "higher frequencies use more power" becomes muddy in situations where the CPU can clock-gate parts of the chip when idle (cpuidle) and where the CPU voltage can change.
It was proven nearly a decade ago that if you don't change voltage at all with clock AND you have a good cpuidle implementation - it is actually best to always clock the CPU at maximum frequency. When voltage changes are in effect - it's harder to tell.
On most devices, the voltage for 500 and 200 are nearly identical. 500 does, I believe, have a somewhat higher bus frequency. So for a given workload, 500 MHz at, say, 20% load will use not much more power than 200 MHz at 50% load. In some cases, a device running at 500 MHz will finish a task more quickly and enter deep sleep faster.
Pretty much - 200 vs. 500 is really questionable in terms of which is best for power consumption. This is why I always set my screen-on minimum to 500.
Any frequency below 200 MHz is pointless as you can't undervolt those frequencies enough compared to 200 to make them have any benefit - in fact in many cases, adding a 100 MHz step is WORSE for battery.
Edit: One thing to note - In Gingerbread, the cpuidle driver was FAR less effective than it is in ICS. Only LPA and IDLE states were enabled by default, and the target residency for both was 40 ms.
In ICS, LPA, AFTR, and IDLE states are enabled and the target residency is 10ms. So it can hit deeper idle states far more often. For example, AFTR isn't as good as LPA - but it's better than dropping all the way to IDLE if you can't enter LPA. This is, in general, why the power consumption when wakelocked is much lower in ICS.
The bad news is that the suspend/resume cycle of the device is longer in ICS, AND cpuidle is totally blocked during suspend/resume - so the suspend/resume cycle eats even more juice than it did before, and it was historically one of the biggest users of power. Eventually I want to try and reduce this consumption.

Thanks for the good explanation mate, it has made things clear for me
Will be keeping my note on minimum of 500MHz aswell as it is a good improvement with no or next to no extra battery drain

[CLOSED]delete

delete

It might have to do with having 8 cores in the system. The problem is that we do not have cooling on our phones BUT: per say you run a game that requires full power of the smaller 4-core processors, and we have 8 cores in total, the game will run evenly on all 8 (hence why you don't notice any lag even though it gets throttled) or maybe a little more towards the bigger 4-cores, they generate too much heat. It's better to spread the load with lower speeds than to keep it running on bigger cores (less cores also) at higher speeds. I believe you can modify your kernel settings to change the throttling temperature if the kernel supports it. I've went back to stock kernel because practically there was no difference when it comes to being a regular user. The cooler the processor is the better hardware life you'll have in the long run, after all, phones don't have sufficient cooling to support such processors at high frequency all the time.
If it throttles later on, let's say 70-ish, you'll complain that the phone heats up like the galaxy phones

Which rom has a good battery life?

Hey i tried many diffrent rom bui i can only get 2.5h SOT. Which kernel do you use and which rom to get the maximum battery life without lag?

Battery<----------------------------------->Performance
They are at opposite ends. If you want maximum battery life then there no way to avoid lag.
You can completely turn off 2 or 3 cores, leaving you with a single or dual core phone;
You can also underclock and undervolt the remaining cores, reducing power consumption even more.
This is very battery friendly (in theory), but, as I said, it will give you big lags.
I added a rom video review to the help thread. You might want to check it out.

obol2 said:
Hey i tried many diffrent rom bui i can only get 2.5h SOT. Which kernel do you use and which rom to get the maximum battery life without lag?
Click to expand...
Click to collapse
Sounds like you really need a new battery
GDReaper said:
Battery<----------------------------------->Performance
They are at opposite ends. If you want maximum battery life then there no way to avoid lag.
You can completely turn off 2 or 3 cores, leaving you with a single or dual core phone;
You can also underclock and undervolt the remaining cores, reducing power consumption even more.
This is very battery friendly (in theory), but, as I said, it will give you big lags.
I added a rom video review to the help thread. You might want to check it out.
Click to expand...
Click to collapse
Don't mean to troll, but can't help it. Turn a quad core cpu into a dual core? Really?!
Erm... ok.

thundastruck said:
Don't mean to troll, but can't help it. Turn a quad core cpu into a dual core? Really?!
Erm... ok.
Click to expand...
Click to collapse
Do you have any better ideas on how to increase battery life?
Cores need power to run. Turning them off is a completely logical option.
You really don't need all the cores for most tasks. You can browse Facebook, or whatever, just fine with only two cores.
I am currently running in dual-core mode with max cpu frequency at 1.2 GHz and it still is smooth.

GDReaper said:
Do you have any better ideas on how to increase battery life?
Cores need power to run. Turning them off is a completely logical option.
You really don't need all the cores for most tasks. You can browse Facebook, or whatever, just fine with only two cores.
I am currently running in dual-core mode with max cpu frequency at 1.2 GHz and it still is smooth.
Click to expand...
Click to collapse
I aldeady underclockt it to 1.2 ghz and undervolting it -625000 uV. Shut off to cores and use intellimm. I also under clock gpu to 320 mhz and under volt it 1 step down. I use zen with 1024 kb. Turn off android logger and zram. Use amplify, power nap, greenify and Max 700 mhz screen off. What I can do more?

obol2 said:
I aldeady underclockt it to 1.2 ghz and undervolting it -625000 uV. Shut off to cores and use intellimm. I also under clock gpu to 320 mhz and under volt it 1 step down. I use zen with 1024 kb. Turn off android logger and zram. Use amplify, power nap, greenify and Max 700 mhz screen off. What I can do more?
Click to expand...
Click to collapse
Nothing.
Test your battery. Or better yet, start thinking of getting a replacement or a new phone. With a 2 hour charge something seriously is wrong with your device.
My 3 year old S4-i9500 still comfortably gets through the day. Bought a brand new original battery 6 months ago. Smooth sailing. No under-volting, shutting off cores, under-clocking or any crap like that. Phone is custom rom'med and correctly so if I may add.

obol2 said:
I aldeady underclockt it to 1.2 ghz and undervolting it -625000 uV. Shut off to cores and use intellimm. I also under clock gpu to 320 mhz and under volt it 1 step down. I use zen with 1024 kb. Turn off android logger and zram. Use amplify, power nap, greenify and Max 700 mhz screen off. What I can do more?
Click to expand...
Click to collapse
lower values for read ahead cache are more battery friendly.
Also, bfq is considered more battery friendly too.

thundastruck said:
Nothing.
Test your battery. Or better yet, start thinking of getting a replacement or a new phone. With a 2 hour charge something seriously is wrong with your device.
My 3 year old S4-i9500 still comfortably gets through the day. Bought a brand new original battery 6 months ago. Smooth sailing. No under-volting, shutting off cores, under-clocking or any crap like that. Phone is custom rom'med and correctly so if I may add.
Click to expand...
Click to collapse
In two month I will get a new phone. I want to buy the lg v10 but don't know if it is worth or lg g4 / moto x style.

obol2 said:
In two month I will get a new phone. I want to buy the lg v10 but don't know if it is worth or lg g4 / moto x style.
Click to expand...
Click to collapse
Never heard of these models except for lg g4, which was crap imo.

thundastruck said:
With a 2 hour charge something seriously is wrong with your device.
My 3 year old S4-i9500 still comfortably gets through the day
Click to expand...
Click to collapse
You are aware that he said 2.5h SOT, "screen on time" ?
That means that after using it for an entire day or two, the combined time of the screen being on is 2.5 hours, and in my opinion that is not really bad.
The screen is always the culprit when it comes to power consumption, and in a phone with a battery that doesnt exeed 2000mah its just natural
that after about 2-3 hours of the screen being on the little battery is empty, and no ammount of undervolting will extend that significantly.
The only thing he can consider to increase SOT is a bigger battery, if he doesnt mind the bigger bulge at the back.

Backe888 said:
You are aware that he said 2.5h SOT, "screen on time" ?
That means that after using it for an entire day or two, the combined time of the screen being on is 2.5 hours, and in my opinion that is not really bad.
The screen is always the culprit when it comes to power consumption, and in a phone with a battery that doesnt exeed 2000mah its just natural
that after about 2-3 hours of the screen being on the little battery is empty, and no ammount of undervolting will extend that significantly.
The only thing he can consider to increase SOT is a bigger battery, if he doesnt mind the bigger bulge at the back.
Click to expand...
Click to collapse
Had no idea - never really cared what SOT stands for anyway! But thanks for clearing that up. Always had satisfactory battery life on my S4. The few times I didn't was when a rom on my phone went bad & subsequently had to flash another.
So if what you say is true regarding a day or 2 battery life, then what is this chump complaining about? Turning water into wine? :silly:

So @GDReaper idea is right!I currently am on crDroid and when I switch off 2 or 3 cores the phone is as fast as it was and the battery lasts a little bit longer..With normal charging it goes to full for about 1 hour and half and without using it the batttery lasts 1-2 days,using it lasts about 3 hours or 4.I flashed a new kernel and turned on fast charging on Kernel Adiutor,now the phone charges full for about 45 minutes and lasts same! I'm very happy with my battery bcs I haven't changed it since I got the phone!

Why is rn5pro performing better than rn7 in terms of gaming performance like pubg ?

I have my friend who has redmi note 5 Pro and his device is performing better than my device which is the redmi note 7 and even the heating issues are more in redmi note 7 when compared to redmi note 5 Pro. I get like 5hrs sot on a full charge for casual use. I'm on miui 10.3.6.0 global stable. Better gaming performance would be my priority right now. A help would be much appreciated

5 hours sot on gaming?

az09512 said:
5 hours sot on gaming?
Click to expand...
Click to collapse
On casual use. While gaming it gives 3 hours and 40 mins

both phones have 4000mah battery and similar FHD+ screen.
the difference is: RN5 has SD636 and RN7 has SD660.
both are fabricated in 14nm but SD660 is more powerful, with better GPU and higher clock speed. that's why RN7 consume more power compared to RN5.
Unless you under-clocked the CPU and GPU, there's no logical way for RN7 to have better battery life compared to RN5; under the same circumstances.
but if we're talking about gaming, yes. RN7 has better graphic performance compared to RN5.

vexa said:
both phones have 4000mah battery and similar FHD+ screen.
the difference is: RN5 has SD636 and RN7 has SD660.
both are fabricated in 14nm but SD660 is more powerful, with better GPU and higher clock speed. that's why RN7 consume more power compared to RN5.
Unless you under-clocked the CPU and GPU, there's no logical way for RN7 to have better battery life compared to RN5; under the same circumstances.
but if we're talking about gaming, yes. RN7 has better graphic performance compared to RN5.
Click to expand...
Click to collapse
Not really true. 660 is designed to be more battery efficient than the older 636. Check the performance comparison provided by Snapdragon.

Because the 660 is clocked at a higher speed and has a faster gpu. I have both phones and you get about 10% to 20% better performance with the 660 but you lose about the same on battery life.
For example the RN5 gets from 8 to 10 hours and the RN7 gets 6 to 8, depending on the games or features you are using of course. Maybe the 710 or 712 have similar performance to the 660 while having the battery life of the 636.

my neighbor has rn7 i have rn5 in free fire mine loads faster than his. i got to say that i use custom kernel and i have franko kernel manager and set profile for free fire to all cores at 1113mhz. and performance is same heat is the same.

kennyk09 said:
Not really true. 660 is designed to be more battery efficient than the older 636. Check the performance comparison provided by Snapdragon.
Click to expand...
Click to collapse
No, it isn't, SD636 is an underclocked SD660 with a weaker GPU. The SD636 came after SD660, not the other way around like you said.
---------- Post added at 03:05 PM ---------- Previous post was at 03:01 PM ----------
Kulbhushan69 said:
I have my friend who has redmi note 5 Pro and his device is performing better than my device which is the redmi note 7 and even the heating issues are more in redmi note 7 when compared to redmi note 5 Pro. I get like 5hrs sot on a full charge for casual use. I'm on miui 10.3.6.0 global stable. Better gaming performance would be my priority right now. A help would be much appreciated
Click to expand...
Click to collapse
The "performance" you mention is only battery related or actual gameplay related too?
If it is the later pay attention to some graphic changes from one device to another, Android games usually tend to add more visual effects according to the SoC being used, which in turn makes the game heavier.