Just a quick question on degredation of a CPU?

newls1

Supreme [H]ardness
Joined
Sep 8, 2003
Messages
4,607
I just have a simple question regarding CPU degradation. Im operating my current 9900K @ 5.07Ghz all core @ 1.285v. This pc sees VERY VERY VERY little hours of operation and its only for gaming, and PC might get 2 hours of play time a week, IF THAT! Do I have anything to worry about in the back of my head?
 
9900K is on the same process as Kaby Lake so I'll leave you with this, and you can interpret the "idle" chart for extrapolated degradation how you see fit. Keep in mind that load and high current naturally will increase degradation rate substantially.

https://www.overclock.net/forum/5-intel-cpus/1673657-cpu-overvolt-death-degradation-stories.html

Let's just say that at that voltage you're running, it's going to be quite a long time ;)
I dont know how i missed that post on that forum, what a crazy detailed reading.. thanks for the link
 
One day in the not so distant future, that overclock will no longer be stable and you will have to nudge it down.

Could be five years. Could be five days.

Agreed... my 6700K was overclocked for a year stable (P95 tested) and then suddenly it became unstable. My FX-8350 has been stable OC'd since 2012.

I decided to stop overclocking altogether... rather have durability than speed...
 
Agreed... my 6700K was overclocked for a year stable (P95 tested) and then suddenly it became unstable. My FX-8350 has been stable OC'd since 2012.

I decided to stop overclocking altogether... rather have durability than speed...

I believe that longevity is much less of a concern with modern chips that have automatic voltage reduction, aggressive idle clock frequency reductions, and the availability of effective aftermarket budget coolers. It is likely only a concern if you're going for a high voltage, hot running, permanent OC (no idle frequency reduction) on the bleeding edge of stability.
 
I believe that longevity is much less of a concern with modern chips that have automatic voltage reduction, aggressive idle clock frequency reductions, and the availability of effective aftermarket budget coolers. It is likely only a concern if you're going for a high voltage, hot running, permanent OC (no idle frequency reduction) on the bleeding edge of stability.

Say what you want but my 6700K became unstable at 1.335 volts after one year which is well under the threshold, so i bumped it incrementally until it got stable which was close to 1.4 volts. By then it wasn't even worth the trade off of higher temps. I've got the Corsair H100i v2 cooler and my CPU very rarely went over 60 degrees at 1.335 volts on P95. At 1.4 volts my temps were hovering around 70 degrees on P95.

My older rig (FX-8350) has been running at 1.5375 volts since 2012 with no issues which is just below its max threshold of 1.55 volts.

IMO, its the luck of the draw and no CPU is immune to degradation. Some OC'd CPUs will degrade quickly, some slowly and some will keep going.
 
I look at things like the 2500k/2600k. Most of those that started to face severe degradation were 5 years+ out from release. At that point, you likely would have bought a new CPU anyway.
 
Say what you want but my 6700K became unstable at 1.335 volts after one year which is well under the threshold, so i bumped it incrementally until it got stable which was close to 1.4 volts. By then it wasn't even worth the trade off of higher temps. I've got the Corsair H100i v2 cooler and my CPU very rarely went over 60 degrees at 1.335 volts on P95. At 1.4 volts my temps were hovering around 70 degrees on P95.

My older rig (FX-8350) has been running at 1.5375 volts since 2012 with no issues which is just below its max threshold of 1.55 volts.

IMO, its the luck of the draw and no CPU is immune to degradation. Some OC'd CPUs will degrade quickly, some slowly and some will keep going.

How long did you test stability with P95 the first time around? Unless you were testing for days on end, it might not have actually been stable initially it just didn't trigger a crash.

I'm dealing with something similar now with memory stability. My memory will pass Passmark memtest and HCI memtest for hours on end, but will crash running certain games inexplicably. Switch out the memory or lower the memory speed and no problems.
 
How long did you test stability with P95 the first time around? Unless you were testing for days on end, it might not have actually been stable initially it just didn't trigger a crash.

I'm dealing with something similar now with memory stability. My memory will pass Passmark memtest and HCI memtest for hours on end, but will crash running certain games inexplicably. Switch out the memory or lower the memory speed and no problems.

I didn't P95 test for days on end...typically 3-4 hours and I was satisfied.

My AMD rig never crashed in 5+ years of OC'd gaming. My 6700K same story except it started to blue screen while gaming after 1 year...
 
My 4770K has been running at 4.4 for almost 5 years now-no issues with crashing etc
 
Last edited:
I've only experience degradation when running very high voltages, I wouldn't worry about it.

Honestly that's anywhere from a 0 to 300mhz OC over factory boost clocks. If you are worried about it, I doubt you will see much difference going back to stock clocks.
 
Last edited:
No. Forget about it. Your OC is so close to the stock turbo boost it's not even worth thinking about.

It’s so close to the stock turbo it’s not even worth doing. Zero chance you’ll notice the performance difference, 100% you’ll notice the heat difference.
 
The motheboard will need replacment before the CPU usually. People think their CPU is degrading, when it's really the motherboard. The only way your CPU will degrade is if you are on some sort of extreme cooling, where it's getting thermal stress from extreme heating and cooling constantly. Or you are running it at thermal max package levels with poor cooling, in which case the package would degrade.
 
Not necessarily. High voltage can also do this. There were plenty of reports of 2nd and 3rd generation chips degrading early at high voltages. 2500Ks and 2600Ks at 1.5V+ even with top-quality cooling might last a year before overclocks need to drop. I never let my 2600K go above 1.38V and used it for almost 8 years at 4.5Ghz, but there was always a chance it could've gone the other way. I've read that electromigration from high voltage is a bigger issue on smaller process nodes but I haven't really looked into it much.
 
Back
Top