Overclocking Stability Over Time? i5 750

[limniscate]

ok my system was stable for 3 months at 3.91Ghz 186 blck, 1.36875 core voltage, imc 1.24, dram 1.66, pll 1.86, pch 1.11, but I've run into the problem again. Intelburn test will BSOD after 1 iteration and Starcraft 2 crashes.

Am I killing my chip? I'm trying to find stability at 3.8Ghz now. At blck 180, core 1.325, imc 1.24, dram 1.64, pll 1.86, pch 1.11 it BSOD on the 30th iteration of intel burn test max settings.

 

[Spikesoldier]

3.7 is a sweet spot voltage-wise i would try 3.7 first and see if that is stable

 

[996GT2]

ok my system was stable for 3 months at 3.91Ghz 186 blck, 1.36875 core voltage, imc 1.24, dram 1.66, pll 1.86, pch 1.11, but I've run into the problem again. Intelburn test will BSOD after 1 iteration and Starcraft 2 crashes.

Am I killing my chip? I'm trying to find stability at 3.8Ghz now. At blck 180, core 1.325, imc 1.24, dram 1.64, pll 1.86, pch 1.11 it BSOD on the 30th iteration of intel burn test max settings.

You're overclocking too high for what your heatsink can handle, especially since you said temps were getting into the >80C range. Either get a better heatsink like a Noctua NH-D14 or lower your overclock and voltages.

 

[limniscate]

You're overclocking too high for what your heatsink can handle, especially since you said temps were getting into the >80C range. Either get a better heatsink like a Noctua NH-D14 or lower your overclock and voltages.

Thanks for the info. My system never actually gets over 80 in actual use though--does that matter? It only gets that hot when I stability-test it with IBT. The most taxing thing I do is StarCraft 2.

I don't think it goes over 60 degrees Celsius in actual use.

 

[podspi]

Try returning everything to stock and see if you are stable there. Whenever I have any hard-to-find issues like this, the first thing I do is put everything back to default. That way, at least you know nothing is critically broken and you don't spend forever messing with settings for no reason.

 

[DirkGently1]

Try returning everything to stock and see if you are stable there. Whenever I have any hard-to-find issues like this, the first thing I do is put everything back to default. That way, at least you know nothing is critically broken and you don't spend forever messing with settings for no reason.

Good advice.

There's a better than even chance that this fault lies with the motherboard rather than the CPU. The only way to know for sure would be start swapping components, but this is impractical if you don't have spares.

Considering the time of year, is your environment getting hotter? Just measuring CPU temps won't tell you if there are hotspots on your mobo causing problems. Perhaps the old trick of having a desktop fan blowing into your PC could help rule this out?

 

[yottabit]

Do you have C1E or Turbomode enabled?

Is your BIOS up to date?

I'm trying to think of how long 17 passes takes. I know I had similar issues as you (same motherboard) and it was RAM that was the culprit. I ended up having to run it at 1333 @ 1.5 volts. I did Memtest86+ for 24 hrs I think... you'll be surprised, sometimes errors won't show up for 10, 12 hrs...

Starcraft 2 crashes are related to memory most of the time IMO

Also think about this: that's a good amount of voltage you're putting through the chip, is it really worth it? I don't think the gains you'll see in SC2 really warrant it. In the realm of 3.5 ghz your GTX460 is probably just as much of a bottleneck as your CPU (if you are playing on Ultra 1080p or something)

 

[hardboy]

Your system well need a better psu sooner or later. It is best to get a psu with 300 400 watts extra so that it lasts 1 or 2 major upgrades

 

[limniscate]

I don't have C1E or Turbo mode enabled.

My BIOS is up to date.

I think it's stable again; it's just frustrating to have to re-stabilize every few months. I ran 40 passes of IBT at max at 3.8ghz, 1.325 vcore, 181 bclk, 1.24 IMC, 1.66 dram, 1.86 pll, 1.11 PCH.

It is hotter now than when I first got the computer since I'm in TX.

 

[Blastman]

Considering the time of year, is your environment getting hotter? Just measuring CPU temps won't tell you if there are hotspots on your mobo causing problems. Perhaps the old trick of having a desktop fan blowing into your PC could help rule this out?

This.

I've seen OC systems in the past that were built during the winter months and running stable, run into problems in the hotter summer months when all of a sudden the motherboard components are running hotter because the temperature in the room where they are situated in is warmer.

 

[BababooeyHTJ]

I had the exact same thing happen with a P45. It seems like something degraded, probably the cpu. Are you using loadline calibration by any chance?