Poppinj

So, if you change the voltage to the cpu, or change its speed on my desktop then the stepping program that throttles my cpu is disabled. Its running full speed all the time. Wouldn't the same thing happen on a laptop? Most of the energy savings from a laptop come from throttling the cpu when not in use so you would need to really be saving a lot of power from your under volt in order for this to be worth while.

Grim

Thats correct.
But what these programs allow you to do, is FURTHUR controll how the machine does that.

You can make it run even slower (when you really dont need the CPU power), AS WELL AS, undervolt the chip, thus saving MORE power.
(NHC has some nice presets btw).

Poppinj

I guess I dont understand how this all works then. If you want a slower power efficent cpu, then dont buy a gaming laptop, buy a laptop with a slow power efficent cpu. It depends on how well the computer throttles the cpu with the reduced voltage I suppose, but I though decreasing voltage would dissable the computers ability to throttle the cpu frequencys...

0rion

Interesting article. I have an IBM laptop that gets about 4 hours of battery life when im listening to music, or doing work on it. It'd be very interesting to see If I could get another hour out of it by undervolting.

Grim

I bet with THAt, and using the NHC presets..


at least 30mins more .

Would be interesting to see some results (more).

MY friend's lappie now lasts 35 or so mins longer.
But thats without undervolting

Skullzaflare

just to add, while on a laptop if trying to save battery life make sure there is no cd in the drive unless your using it as windows likes to randomly spin the drive up
also alienware laptops have a cmos reset

__________________
Miata joking lists: "You Know YOU are a NA Owner when..."
ive done it lol

Suigi

I swear by RMClock - it really is a wonderful little tool for laptops.
I've tried to undervolt my little lappy a few times already. I was able to take a full quarter of a volt off the highest speed's voltage setting, but it was unstable.
I guess I could stand to try it again - my Core Duo T2300E can push that 70 degree limit when both cores are loaded to the max for a while, so when I do, I'll post my stable results here.

@Krunk: Celerons don't have SpeedStep, so you can't change your CPU speeds. You should be still be able to undervolt, tho.

Krunk_Kracker

Ahh, ok....well, it wont hold the undervolt though either

__________________

Suigi

That's unfortunate, but I guess what you get with Celerons.
I've dug up my old undervolt settings, and where I'll be starting this time around.
Keep in mind that 0.950V is NOT the minimum that a Core Duo can be set to, but due to on-chip restrictions, it can't go below that through software.
Multiplier Result
6x 0.950 MIN
7x 0.950 MIN
8x 0.950 MIN
9x 0.962 STABLE
10x 1.262 stock, 1.025 STABLE under 9hrs Orthos (but I do remember it freezing under regular usage - might've been a CxE state issue)

Grim

Nice that you should mention that.

An important thing to remember is that CPU's FSB's only need a certain amount of voltage to operate..
multipliers are something that came into being in the latter days of the CPU evolutions, and it is the MULTIPLIER, that draws the majority of additional voltage that we overclockers give to our chips to achieve stability.

Consequently, if one strives for an OC that is more based on a higher FSB, rather than a raised multi - then stability can be acheived with relatively lower voltages (try it on your OC ^_^).

Anywho - the same is true for undervolting.

the lesser the multiplier, the less voltage is needed.