Is it bad to decrease the voltage?

[MisterDuck]

I'm running a really small lan box, and there's just not a lot of room around there for cooling or airflow. I noticed that the cpu was running a bit hot, so I decided to try decreasing the cpu voltage a bit to see if that helped - it did (temps went from average of 50c under load to about 45c under load) and I didn't notice any stability issues (I ran the box at 100% cpu use for about an hour to see if I could make it crash).

What I'm wondering is: can I damage the CPU by decreasing the voltage? I would assume that you can only damage it by increasing it and the worst that will happen by decreasing it is that it won't boot up or be unstable, but I just decided that perhaps a second opinion was in order....

In short, does anyone know if you can damage your processor doing this? It's currently set at 1.675 volts, and I believe the standard for the processor (athlon xp 1500+) is 1.7 - think I could drop it even farther? I'm not interested in oc'ing the comp - I just want it to run stable and cool.

 

[CTho9305]

the lower the better. if you go too low, it will start to crash - just bump it back up. running on a lower voltage is better.

 

[AndyCunningham]

You'll get lower power usage as well from lower voltages. This is the same thing that Intel does with mobile chips anyway, so it's perfectly safe - some of them run on I think 1.1V

 

[dunkster]

Any VCore above what's necessary for stability just increases cpu temp.

I run XP1600@148MHz FSB game-stable at 1.775V, 41C loaded temp (I haven't much choice, since this XP1600 is unstable with VCore over 1.8V regardless of FSB clock speed).

Hope this helps!

 

[MisterDuck]

So nothing besides general instability will result from me decreasing the voltage even a bit more? I just want to make sure I'm not going to damage anything...

Currently, the MBM reads cpu core voltage as being 1.67 - does anyone have any idea how low I could theoretically go before encountering stability issues?

 

[Wolfcastle]

I was at 1.65v, and it ran ok. It pretty much in the reverse direction of overclocking. When you overclock, you just keep upping the speed until you run into instability. Then you bump up the voltage a bit. So when you lower the voltage, it's almost like you started from a lower speed and and lower voltage and kept increasing the speed till you got to stock speed. Just keep lowering it till you find a point where it's no longer stable, and then raise it again to its stable point.

I've never heard of a CPU being destroyed simply because the voltage went too low - as long as you do it incrementally, of course.

 

[CapsLock]

I'am doing the sam ewith my celerom 900, lowering the vcore. if you consider U=R*I and P=U*I you get P= u2/r so lowering the
voltage will decrease the power used by the CPU significantly. the main thing is that at some point while lowering Vcore some
transistors on the die will not switch fast enough and the cpu will crash. but crashing was not killed any Cpu yet, as far a I know.
(how could you use Windoze if it had....?) What kills Cpu's is heat. the silicon can stand a certain temperature, depending on
manufacturing process. that's why you cool the cpu ;-).

cAPSlOCK

 

[Jhhnn]

Yeh, just lower the voltage a notch at a time til you get instability, then bring it back up to where it runs right. Repeated crashes can lead to registry corruption, so don't let it crash too may times in a row.... What you're doing really can't hurt the hardware, though.