Overclocking "Drift" - 3.6 P4 Slowing Down ?

[Raviolissimo]

Asus has this synthesized voice "System Failed due to OC'ing"

when i first got this thing, it OC'ed GREAT (well, if you call 11.1% great) for about a year.

specs -
Asus P5GD
Intel 560, 3.6 P4
Patriot DDR2-533 low latency, 3-2-2-8
Maxtor 300 GB SATA
Asus X850XT

for about a year, it was rock solid stable with an 11.1% OC.

it was Anandtech's original article that got me pysched about it
http://www.anandtech.com/showdoc.aspx?i=2293&p=33

their best result was 227 FSB, no big deal, i know, "Big Whoop".

- - -

i admit 11.1% isn't much of an OC.

but now the thing won't OC at all ! or, rather, i'm getting frequent enough error messages that i just went back to stock speeds.

has anybody ever had this happen, where a system runs OC'ed fine for a year, then at some point starts to fail OC & has to return to stock speeds ?

 

[gomerpile]

take a look at the coils on the mb is the copper darkend at all, I can say I had the same problem with my xp system. What I did was get a new set of coils and installed them very risky but well worth the time spent.
http://www.excesssolutions.com/cgi-bin/item/ES2208
This is a pic to show you what you will be looking at some will look different.
Its is the coils that supply the cpu with power.

 

[wun911]

dust builds up over a year....

 

[gomerpile]

Why Did I not think of that I always go for the worst I suppose that is what makes me me. Running around getten killed alot.

 

[Raviolissimo]

minimal dust, inductors/transformers look normal.

i run it with case covers off.

also, the room it's in now is cooler temperature wise than where it was when i first got it.

 

[dimwhited]

what was the vcore when you were overclocking???

if it was too high you cpu could have been damaged, and not be able to go as fast anymore

 

[Raviolissimo]

the whole configuration was pretty close to stock.

to OC it i increased the memory bus speed from
200 to 222, when i first got the system.

voltages were default.

 

[dimwhited]

I would check cpu-z to see what it says the voltages are. just because you have it set at something, doesnt mean thats what it is.

Its possible that a bad mobo or psu (probably mobo) could have caused the ammount of voltage being giving to the cpu to decrease, which would make it unstable

also, i would just try upping the vcore on you cpu a few notches to see if that makes any difference (but if your doing it in a software OC program, dont do it all at once. Go up one notch, then wait a few seconds, then go up annother. Software OC programs have a tendancy to crash if you take to big of steps (atleast mine always does)

 

[Raviolissimo]

actually, the system works fine at 3.6. one of the re-boots/system hangs cost me about a half hour, which is about my limit on tweaking. i'd rather learn 3D Max !

i'm just curious if other people have experienced the same thing - a system that looks rock-solid, fans still spinning & cooling, changing over time. i don't mean a hard drive crashing, i mean a 10% change in performance.

it's like the system is getting older - but this is silicon ... i thought silicon was pretty inert as long as it was properly packaged (cooled).

 

[ParoXoN]

Silicon is typically pretty inert, but there's no guarantee that things like metal or cap electrolyte will stay in their originally-intended condition.

Had you noticed any temp changes before the crashed started occurring? It could be that your thermal compound has baked to the point of ineffectiveness (though that strikes me as unlikely.)

It could just be long term stress on the CPU, as well. Processors do have a running life and OCing will typically reduce that.