Can OCs 'burn in'?

[kingdingeling]

My main rig has been running almost 24/7 since December, OCed to 400x8 = 3.2GHz on my Gigabyte 965P-DS4. The only changes I've had with this system since then were

1.) Swapped the X1950Pro for a 2900XT

2.) Because of the 2900XT swapped my 500W BeQuiet PSU for a Enermax Noisetaker II 600W

3.) swapped my 2GB GeIL PC2-6400 (Elpidia chips) for some Mushkin HP2-6400 (D9GMH).

 

1 and 2 should not have any effect on my CPU/RAM OC, and the Mushkin's behave pretty much the same at 400MHz as the GeIL's (4-4-4-12 on both sets). These settings have never failed me under any circumstances. But just 30 minutes ago, I was playing a little bit with my E6400 since I got a Thermalright Ultra 120 eXtreme (ca. 8

[iKillSteal]

If 450*8 worked with the Geil sticks and not your new Mushkin sticks then it's possible that you've got a lemon, or a pair thereof. No offense but isn't it a tad bit silly to assume that two different parts will be able to do the exact same speed/timings? As for the OC burning it, I suppose it would be possible that your CPU's stability at higher speeds has degraded because you ran it OC'd like that for almost a year...Don't have any experience with something like that so I really can't say.

[cchalogamer]

I've noticed some strange things with OCs in the past, and by nature it's probably more of a luck type thing than any real "trend" but here goes anyway. Based on MY EXPERENCE AND GENERALIZATIONS OF WHAT I HAVE EXPERENCED AND WE'RE TALKING MY STUFF ALONE HERE:

 

Skt 478 Northwoods and 939 Winchesters/Dual cores all seem to gain some form of stability over time when running at near the max speeds. Especially notable with my old 3000+ winchester, If run 24/7 on air cooling @ 2.5ghz (which needed heafty voltage for that one putting me on the edge of my thermal comfort zone for winchesters) About a month later it would easily run lower voltage with a 2.6ghz OC though it would slowly loose that stabiltiy and fall back to 2.5ish assuming the cooling didnt change in under a week. THe opposite was true for my 3000+ Venice, it liked to get there hard and fast, but without a few "breaks" in the higher voltage/speed action it would start loosing some stability around the 2.8 mark, all of which could be gained again by dropping back to 2.6 or so for a few days.

 

Moving to the 775 world I havent had as much long term testing here, but the E4300s ive had have all been of the burnin time = more stable OC type, yet my E6600 doesnt like sitting above 3.7 for very long without a little "extra" cooling but after backing to 3.4 or so for a few days It'll be rock solid over 3.7 for a few weeks. Of the 3 E4300s the two I spent real time with seem to follow the gained stability over time path. The E6400/6300 I had/have dont seem to fit anywhere into this chaos. the E6400 seemed to stay steady at the max OC i had it set to and never seemed to gain from burn in, however It was never in a "great" Ocing board for CPUs with multipliers under 9 in my hands (now sits in a friend's eVGA 680i T1 like my E6600) so it may not really count here. As for the E6300 I have, screw it, it can be the most frustraiting CPU Ever at times. I actually thought it was this damn MSI board until I tried an E4300 in it and found out it was the 6300, not the board, being "picky" about something.

 

/end total sillyness

 

So yeah If you cant figure out what the hell that was all about dont feel bad, I'm not sure either. It's just to show that CPUs do some really wierd crap at times when OCed and dont always do as you would expect. So I guess it IS possible your CPU/some other hardware has become acustomed to a certain speed and functons better at it now. It defies logic but when I start talking overclocking it's the one time this programmer tells logic to take a vacation. To me Ocing is like a trip to vegas, only this game cost me less...I think (You know now that I let logic into this conversation it tells me all these CPUs, motherboards, cooling devices, etc might not be as expensive as a week in Vegas for me...)

[kingdingeling]

If 450*8 worked with the Geil sticks and not your new Mushkin sticks then it's possible that you've got a lemon, or a pair thereof. No offense but isn't it a tad bit silly to assume that two different parts will be able to do the exact same speed/timings? As for the OC burning it, I suppose it would be possible that your CPU's stability at higher speeds has degraded because you ran it OC'd like that for almost a year...Don't have any experience with something like that so I really can't say.

I'm thinking I got a lemon as well... As for the two parts doing the exact same speeds and timings, the GeIL's are rated at 4-4-4-12 @ 400MHz, and the Mushkins 4-5-4-11 @ 400MHz. I was running both of them at 4-4-4-12 @ 400MHz, the GeIL's for like 6 months, and the Mushkin's for 3 I think. The GeIL's would do 450MHz @ 4-4-4-12, but the Mushkin's wouldn't even boot 450MHz @ 5-5-5-15..... So I am a little pissed <_<

 

As for your response cchalogamer, I got the jist of it I think With all three 939 CPU's I've owned, the stability didn't degrade... but I can't really judge that, as I was only running them part time OCed, the longest would have been my X2 4400+, which ran about 8 weeks at 2.7GHz.

 

Anyways... when I have the nerve to try again, the Mushkin's will fly outta this rig. Darn you Micron memory! *shakes fist*

[iKillSteal]

I'm thinking I got a lemon as well... As for the two parts doing the exact same speeds and timings, the GeIL's are rated at 4-4-4-12 @ 400MHz, and the Mushkins 4-5-4-11 @ 400MHz. I was running both of them at 4-4-4-12 @ 400MHz, the GeIL's for like 6 months, and the Mushkin's for 3 I think. The GeIL's would do 450MHz @ 4-4-4-12, but the Mushkin's wouldn't even boot 450MHz @ 5-5-5-15..... So I am a little pissed <_<

 

As for your response cchalogamer, I got the jist of it I think With all three 939 CPU's I've owned, the stability didn't degrade... but I can't really judge that, as I was only running them part time OCed, the longest would have been my X2 4400+, which ran about 8 weeks at 2.7GHz.

 

Anyways... when I have the nerve to try again, the Mushkin's will fly outta this rig. Darn you Micron memory! *shakes fist*

Well that sucks, man. Throw them Elpidia's back in there and do some testing to make sure...It's a little sad seeing good stick turn out to be lemons.

[kingdingeling]

exactly, I mean I didn't go and spend 130

[swifty11212]

exactly, I mean I didn't go and spend 130

[kingdingeling]

These are Micron D9GMH.... crappers, the RAMlist also say they could be ProMOS, in which case they should be called ProPOS... <_< Crap, maybe I'll see if I can RMA, although it is somewhat unfair...