• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Instability, eventually...

Nebbers

Nebbers

Senior member
I'm not sure if any real help can be given without me pasting every setting in my BIOS, but maybe there's a particular starting point for figuring out this kind of problem you guys can point me toward.

Running an ASRock Z68 with an i5 2500k with 8GB Patriot Signature RAM, Onboard/Onchip video.

I've successfully overclocked up to 4.6ghz and it was stable running stress tests in Prime95 for ~3 days. My temps were all way lower than the danger zone, so I thought I had this good to go.

However, after a long period of time - say, a week or two - my computer will BSOD. Seems to consistently happen at least every couple of weeks if I don't restart at some point. Now, I could just keep restarting more often but that doesn't really fix the problem.

Is there any setting this is most likely to be - one of the voltages needs to be increased, perhaps? I can definitely post all my BIOS settings if this doesn't really indicate anything.

Just seemed strange to me that heavy loads don't seem to kill it but long periods of time with average loads do.
 
Have you done any troubleshooting yet? Lowering your overclock? Raising voltage? Is your ram overclocked?
 
Run IntelBurn test. Prime95 doesn't seem to stress as much. That OC is high without some serious tweaking. What kind of CPU cooling do you use? With you iGPU enabled plus your CPU OC'd I'd bet the heat buildup is too much. I OC my 2500ks to 4532 (103 x 44) AND I disable the onboard graphics. I use very good coolers with an extra fan. I push for a higher OC and then back off some for stablity. I always run various tests including 10 runs of IntelBurn.

My gut? Back off .1Ghz to 4.5. Remember that the I2500k is 3.3Ghz with 3.7Turbo. If you are at 4.6 Ghz you are really pushing the envelope.
 
Last edited:
RAM is running as normal for these purposes, though I did have a bit of an overclock on it. Thought I should leave it until this is resolved. It happens with RAM running at stock too.

I was at 4.7 originally, and have dropped down to 4.5 with no change in the symptoms.

I will do a reboot and write down all my voltages and such but I have raised the CPU voltage pretty high... close to the highest recommended by a guide I'd read. HTT I haven't played with as much.

Maybe I am expecting too high of a clock here because the temperatures are so low - maybe the chip just can't handle quite that much?
 
guskline said:
Run IntelBurn test. Prime95 doesn't seem to stress as much. That OC is high without some serious tweaking. What kind of CPU cooling do you use? With you iGPU enabled plus your CPU OC'd I'd bet the heat buildup is too much. I OC my 2500ks to 4532 (103 x 44) AND I disable the onboard graphics. I use very good coolers with an extra fan. I push for a higher OC and then back off some for stablity. I always run various tests including 10 runs of IntelBurn.

My gut? Back off .1Ghz to 4.5. Remember that the I2500k is 3.3Ghz with 3.7Turbo. If you are at 4.6 Ghz you are really pushing the envelope.
Click to expand...

You know, I didn't realize that - the onboard graphics are probably a big factor here.

I don't recall the name of the HS/F I have, it's a Cooler Master something. I could look it up but the CPU temp has always seemed way lower than the problem range. The other temps, though, I am not so sure about - it's possible those are shooting up when I'm doing a lot of graphics-intensive stuff.

I'll drop it down a bit and see if that takes care of it - and maybe try getting it back up there when I buy a video card.
 
That's a good cooler. I'm using variations of it on a 2500k and an AMD 1100T. I add a second fan and it really helps drop temps. I own a licensed version of Aida64 which does a very good job of monitoring temps. The free version probably does also. Install the free version and run it to monitor temps. Also the SandyBridge chips don't need as much voltage as before. BE CAREFUL that you haven't over volted. Bottom line is that heat/high temps cause problems. Personally, unless you need the onboard graphics I would disable it.
 
guskline said:
That's a good cooler. I'm using variations of it on a 2500k and an AMD 1100T. I add a second fan and it really helps drop temps. I own a licensed version of Aida64 which does a very good job of monitoring temps. The free version probably does also. Install the free version and run it to monitor temps. Also the SandyBridge chips don't need as much voltage as before. BE CAREFUL that you haven't over volted. Bottom line is that heat/high temps cause problems. Personally, unless you need the onboard graphics I would disable it.
Click to expand...

I'll check that software out, thanks.

Kinda stuck with the onboard graphics for a while, until I eventually get a card... I have a video card from my last PC, but it's like a 7600GT (PCI-E) or something. Putting that in seems kind of ridiculous 🙂
 
When does it happen? At idle, at load? If it is a random occurrence, I would test the ram. Whenever I have random problems RAM is the first thing I test. I would run a memtest86 overnight.
 
I would personally do a part by part diagnosis just to eliminate possible culprits, start off with zero overclock, test the chip overnight w. prime95, then test the ram with memtest, etc. then OC just one part at a time, leave all else at default setting, test for stability, this process has helped me isolate problems before, it works.

key idea, isolated testing, leave all other parts at default setting.
 
nyker96 said:
I would personally do a part by part diagnosis just to eliminate possible culprits, start off with zero overclock, test the chip overnight w. prime95, then test the ram with memtest, etc. then OC just one part at a time, leave all else at default setting, test for stability, this process has helped me isolate problems before, it works.

key idea, isolated testing, leave all other parts at default setting.
Click to expand...

This... I know its a PITA and time consuming but it's really the best way. Taking shortcuts will likely take longer unless you get lucky, becuase eventually you'll give up and go through this processes anyway.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top