O/C Performance of system better when new? Degrades in time???

frodox

frodox

frodox

Associate
Joined
5 Aug 2008
Posts
11
P5E3 Premium (Winbond W83627DHG)
BIOS ver 0402
Q9450 + ThermalRight Extreme
OCZ DDR3 2000 MHZ (OCZ3P20004GK) @ 9-9-9-24

Got this system up to 430x8 (3440Mhz) prime stable overnight.

Good result, was pleased with it. The system was then brand-new.

However, a few months has passed and, over time, have started getting bluescreens and other random stuff. Ran Prime today immediately crashed on two cores.

Seems the performance has degraded since the system was new. I would suppose this would be due to capacitors drying out on the MB? I build the system in summer and the problems have recently arose in November when in UK the winter air humidity has dried where the box is down to 20% relative humidity - very dry air.

Is this timeline/seasonal degradation normal and is there any kind of timeline when a system could be considered bedded-in?

Cheers
 
Last edited:
It's normal for a chip to degrade, so higher voltage is needed for the same speed, but not at that kind of speed, and the performance shouldn't drop. It's certainly not normal, something is obviously defective.
 
Tried flashing the bios to a more recent one?

By degrading do you mean performance in games isn't as good? Encoding/folding etc?

Positive theres not any rogue drivers or dodgy files causing havoc?
 
Generally overclocks get more stable during winter due to the cooler ambient temperatures.

Many people actually reduce their overclocks for summer.

I would assume you can save your overclock settings in the BIOS somewhere (My P5Q Pro has Asus O.C. Profile). Then set everything back to normal speed and then Prime and see if you still get BSOD's. If they continue it's obviously not your overclock and you can start searching elsewhere. If they stop then reload your overclock and try backing it down until they stop :)
 
I'd test your memory.

You shouldn't have had to go anywhere near dangerous voltage levels for a Q9450 when running it at 3.44ghz.
 
frodox said:
Please explain this? Do you refer to the CPU alone?
Click to expand...

It happens to all everything that uses semi-conductors I believe, it's just normal reactions in the chips, sped up a bit by the higher temperatures but it won't be what's happening here, it takes years not months to make a significant effect in a CPU.
 
You must log in or register to reply here.
Back
Top Bottom