SNDS

[TAKKLE]

Just wondering here, what voltage should I stop at to aviod SNDS? I think I remember hearing 1.65v was about the highest you could go.

Also, will a water block for my NB help any here, is it worth it to get one if I have a HS and fan on mine?

 

[Soulkeeper]

well it all depends what cpu you are using and what process it is made with (the transistor gate size ie: .18u .13u or .09u)

Most people in these forums seem to agree that if you go over 1.7v (a little less maybe 1.68) with a P4 (northwood core .13u) you will damage it

the athlon (barton and t-bred cores .13u) seem to take a little bit more voltage before they die on you at around 1.74 to 1.76v in my experience, although many seem to enjoy running them higher

the P3 (tualatin .13u) should be somwhere around these limits as well

the jury is still out on how much voltage the new .09u processors from both amd and intel will be able to handle but i'm will to bet it will be atleast .1v less

as far as older .18u processors are concerned there are many and each is a bit different
athlon thunderbird seemed to run fine up to 1.8v and even higher (some people ran them over 2v, although this never seemed safe to me or the athlons i destroyed)

the P3 (.18u) i'm not too sure what voltages would kill it but probably around the same ballpark as the athlon

the P4 (willamette .18u) would prob take as much as 1.8v but i'm not too sure cause i never owned one

remember tho a lot of this is all debatable and no two processor are exactely the same and the death of a processor could take minutes weeks or even years when overvolted


i know i left out a lot of other processors but i hope this helps
btw i've destroyed atleast 5 athlons by now from overvolting
nomatter what kind of cooling you use voltage can still kill them
believe me i was using subzero watercooling and still managed to destroy many

what do you plan to overclock anyways ??



i guess someone else can correct any errors or fill in the gaps
they could prob explain the whole electron migration thing better too

 

[jjyiz28]

soulkeep nice, most people have misconception that as long as your cpu core is cool, then voltages like 1.9, 2.0 cannot hurt you. not true, even for athlon chips.

 

[TAKKLE]

Well you explained everything I needed to know, thanks

I'm going to be OCing a 3.0c P4. I'm trying to hit 4 Ghz with 1:1 timings, and my 2.8 just won't do it lol we'll see how it goes here next week hopefully

 

[TAKKLE]

Originally posted by: jjyiz28
soulkeep nice, most people have misconception that as long as your cpu core is cool, then voltages like 1.9, 2.0 cannot hurt you. not true, even for athlon chips.

Exactly why I posted here, I thought SNDS was because of heat on the chipset, but suspected otherwise, now I know better

Thanks

 

[Soulkeeper]

one way i like to think about it is by comparing a cpu to a lightbulb
you can only run so much voltage/amps through it
the thinner the wire in the lightbulb the less current you can run through it before the bulb wears out on you

and just like a light bulb the cpu will eventually wear out even at it's standard voltage (several years tho), but the lifespan is inversly proportional to the voltage being used (perhaps at a calculatable exponential factor)

also one thing i've often thought about was that the time the cpu is stressed the most is when the computer is turned off and on,
also like a light bulb cause that is usually when they burn out

 

[TAKKLE]

Aye, I don't remember the last time I had a light bulb burn out on me except when I turned the lights on lol

Good analogy.

 

[Shimmishim]

Originally posted by: TAKKLE
Well you explained everything I needed to know, thanks

I'm going to be OCing a 3.0c P4. I'm trying to hit 4 Ghz with 1:1 timings, and my 2.8 just won't do it lol we'll see how it goes here next week hopefully

1:1 is going to be tough... then again i was trying to run 2-2-2-6 timings at 250 mhz

enjoy the cpu when it comes... it's an awesome chip! sad to see it part but i have a new project coming up...