Windows randomly hangs on OC

[gumpty]

Hi guys, hopefully I can get a bit of help with the problem I'm having overclocking my i5-760.

So, basically the problem is that no matter what overclock I put into my CPU, I've had it stable through hours of stress tests at 4Ghz or 3.8Ghz, for some reason it will randomly hang in Windows when doing just plain innocuous things like surfing the net or doing nothing at all. It happens randomly, it might happen after 30 mins, or it might happen after 2 hours, but it always seems to happen.

I'm really at a loss. It's not like it's an unstable* overclock - it runs stress tests fine and I've never had it over 65ºC and never had a BSOD. It's essentially a fresh Win 7 install too.

I've disabled Turbo Boost, but suspect it may have something to do with the other power states and options as in this screenshot:

But running a overclock with these disabled still hangs. Maybe I should be running it with C1E & C3 etc enabled?

Thanks for you help.

* Of course it technically is an unstable overclock, but whatever.

 

[hat]

Disable C1E, EIST, C3/C6/C7 State Support (not sure what this one is, but it sounds like a power saving feature).

That sounds like a very strange issue. What stress test did you run? I prefer 6 hours OCCT Linpack. You should make yourself a bootable Memtest86+ disc and let that run overnight to give your memory a real thorough screening. Check your GPU oc too, set it to stock and see if this problem still happens. GPUs can cause many a strange thing to happen...

 

[micropage7]

what clock you are now? if its kinda high maybe you need raise the processor voltage
what about ram clock, check the clock too or you need to set it manually

 

[gumpty]

Disable C1E, EIST, C3/C6/C7 State Support (not sure what this one is, but it sounds like a power saving feature).

Yep, done that, still hangs. There might still be some combination of 'enabled/disabled' of these functions that I haven't tried.

That sounds like a very strange issue. What stress test did you run? I prefer 6 hours OCCT Linpack. You should make yourself a bootable Memtest86+ disc and let that run overnight to give your memory a real thorough screening. Check your GPU oc too, set it to stock and see if this problem still happens. GPUs can cause many a strange thing to happen...

I've run Prime95 and Everest Ultimate Aida64, probably should throw others at it to be sure. Can't run them overnight though, as my PC is in my bedroom and having it on overnight would just give my fiance another reason to hate it. Might try it while I'm at work.

GPU is at stock.

I do wonder about my memory - it was the cheapest that I could find. I have heard that quality ram is essential for a decent overclock, but I have these issues even at a tame overclock of 3.4GHz. I do intend to get some quality ram some day, but I want a 2x4GB set with decent timings. These are pretty hard to find at sensible prices at the moment. Besides, I would have thought a memory problem would lead to a BSOD, rather than a freeze? (I don't know why I think this, just intuition)

what clock you are now? if its kinda high maybe you need raise the processor voltage
what about ram clock, check the clock too or you need to set it manually

I'm at stock now cause I'm sick of it hanging.

I bumped the voltages to get 3.8+GHz to the following:

• Vcore = 1.35
• QPI/VTT = 1.35
• PCH = 1.14
• CPU PLL = 1.9

I've used the stock XMP profile for the ram, but dropped the multiplier thingy so it was never above 1600MHz.

 

[Mussels]

i had issues like yours from VRM/NB/memory overheating. they dont have temp sensors so its hard to diagnose, and the heat takes hours (or days) to build up, so it happens at random times.

do a finger test and see if any of those are too hot to touch, at various times (right after a crash, is the best time)

 

[brandonwh64]

Try Intel burn test 2.5. put it on max memory and 4 threads (since you dont have HT)

http://www.google.com/url?sa=t&source=web&cd=3&ved=0CCkQFjAC&url=http%3A%2F%2Fmajorgeeks.com%2FIntelBurnTest_d5987.html&rct=j&q=intelburntest%202.5&ei=Dxz-TPfqHcT_lgeg8r3FCA&usg=AFQjCNGR3nRysDPa5O8ZN-pVLKb3iT1T3Q&cad=rja

 

[Bundy]

Random unexplained crashes are classic signs of RAM instability. As Mussels suggests, check your temperatures. I also suggest you check (I''ve never used an i7 so I'm typing out my arse here) that the RAM volts are ok.

If the above issues are fine, it leaves the dissapointing news for you. As the memory vcontroller is on the CPU in this processor, it also seems likely that your OC is not really as stable as you would think. If the problem goes away when you slow the CPU down, then your OC is at fault - as what hat and micropage are suggesting.

 

[brandonwh64]

Also try memtest86+ for about 3 or so hours to see if it spits any errors back at you, that would help as well

 

[gumpty]

Hmmm ... excellent, thanks for the tips guys.

Yeah, I have wondered how the ram would be coping since I've not got direct airflow over them with the H-50 further away. Also another thing I suspect might be an issue could be that this motherboard has no heatsinks or cooling on the mosfet/VRM thingies.

Will have a play around when I get home tonight.

What brand of RAM would you guys recommend, if I was to replace it.

 

[Mussels]

Hmmm ... excellent, thanks for the tips guys.

Yeah, I have wondered how the ram would be coping since I've not got direct airflow over them with the H-50 further away. Also another thing I suspect might be an issue could be that this motherboard has no heatsinks or cooling on the mosfet/VRM thingies.

http://images.bit-tech.net/content_images/2010/03/gigabyte-ga-p55m-ud2-motherboard-review/2.jpg

Will have a play around when I get home tonight.

What brand of RAM would you guys recommend, if I was to replace it.

brand doesnt matter, its what chips are on them. if they're overheating the lower the voltages the better, but airflow is the best solution.