Loss of OC stability over time

[jtvang125]

When I first built my rig about a year ago I got my 5820k running stable at 4.5ghz in testing and games. Recently all games started hard crashing after a few minutes of playing. I checked temps and they were within range. I ran tests for twice as long as the time it usually crashes in games but they won't cause a crash, just games. I then drop my oc to 4ghz and I can now play any game for hours on end.

What happened? Is there anything I can do to get it back to 4.5 and be stable?

 

[Smoblikat]

Is this on an ASRock board by chance? Ive noticed that their VRM's start going bad really quickly after a certain point. Love the boards though, they just dont age well.

 

[HutchinsonJC]

As a basic concept, it's generally pretty well accepted that an overclock, over time, will become increasingly less stable.

The harder you push a cpu, the more dramatic that drop off in stability will likely be.

There's a whole assortment of variables that come in to play when looking at stability, and as such, there can be a lot of trial and error in those various variables to see what works.

I think for anyone to be able to make reasonable guesses, we'd probably all like to know what kind of system you're running (Motherboard, RAM, etc), and what types of voltages you were pushing, and if the OC is a base clock manipulation, multiplier or combo of the two. Were you using built-in overclock configurations or did you tune yourself?

The more you give us, the more a collective us can dissect the details. Otherwise, my first sentence of this post is about as good as it gets.

 

[jtvang125]

Is this on an ASRock board by chance? Ive noticed that their VRM's start going bad really quickly after a certain point. Love the boards though, they just dont age well.

No it's a gigabyte board. So the tolerance of the VRM could have shifted?

 

[Smoblikat]

No it's a gigabyte board. So the tolerance of the VRM could have shifted?

Defintely, especially if the board was lower quality to begin with (most gigabyte boards are from my experience). I recommend posting all the info JC asked above, that will give us a lot more information to work with to try and narrow this down.

 

[Rifter]

My last build before my current build was a i7920@4.2Ghz. It started off at 1.275v, and 7 years later was up to 1.4v to maintain the same clocks, every year or so id get the occasional crash and id up the voltage a tick or two to compensate then it would be crash free for a while then rinse/repeat all the way up to 1.4v at the end. That system is now back to stock clocks/voltage and given away to a friend, who uses it daily, still on the same mobo.

 

[Excessi0n]

I actually recently made a post on OCN about my findings after running my 6700K at 1.5 volts for a year. It was still rock-solid in everyday use, but it was no longer stable at the same clocks and voltages in very heavy AVX loads (Prime95, linpack).

 

[DigDog]

it's an easy enough test; raise slightly the voltages, if it stabilizes, there you have it. the same happened to my gigabyte DS3 board, the only board i've had long enough to see it fail. well, fail, to lose voltage.

 

[DrMrLordX]

OP, did your temps changer at all with that 4.5 GHz OC before it went unstable?