Good Overclocking Gone Bad; What Happened?

[InterMurph]

In this thread, I detailed my overclocking efforts with a Core 2 Extreem X6800. I settled in at about 3.6 GHz. Everything worked fine for several months.

Now my system is spontaneously rebooting. Sometimes it lasts an hour, sometimes a minute.

Edit: it's not rebooting, it's freezing up.

I have backed everything off to the defaults (2.93 GHz, default memory timings, etc.), but it is still happening.

It is actually cooler in my office now than it has been for the past four months.

I can't figure out what the problem is. How can I start?

Thanks.

Update: I moved the video card from its original PCI-X slot (adjacent to the NB chip) to the other PCI-X slot (farther away), and the problem went away. No explanation for why it worked fine for 4 months in the original slot.

 

[Markfw]

Power supply getting weak ? Dust in the case making the PSU hot ? My dual Opteron 248's started re-booting (non-OC'ed) after 3 years of 24/7, I shut down, blew a lot of dust out of everything, and now it seems to be fine. Not sure if it just needed a break, or the dust had an effect.

 

[crispy2010]

I would guess a dusty psu!

 

[3chordcharlie]

Sure sounds like PSU - clean it (and your other fans) and if it still happens, test it or just swap it.

 

[InterMurph]

I opened up the case, and the power supply is almost dust-free. And so are the case fans. I cleaned them off anyway.

I should say that the system is not "spontaneously rebooting", as I said in the original post. Instead, it is freezing up, with the display intact. The screen blanks for less than a second, then the display returns, and all is dead.

Also, last night I ran MemTest86+ on it. It ran for 9 hours with no errors before I got the freeze-up. Then I ran it again, and it froze up after 18 minutes, again with no errors.

This P4 Northwood is killing me!

 

[Roguestar]

Sounds like a graphics card overheat to me. Cleaned out the graphics card heatsink and fan? Check temperatures?

 

[InterMurph]

It's not just the graphics that are locking up. The entire system is freezing up.

The display remains on the screen.

I'll look at the graphics card, but it's not a powerful one at all (Radeon X1300), and it's not doing much (I'm just browsing and buliding cod, no games).

 

[3chordcharlie]

You could be having a driver issue, did you make any software changes?

 

[InterMurph]

No, no software changes.

 

[GuitarDaddy]

In my experiences, freezes like you describe are almost always ram related. And the fact that it occured twice while running memtest just strenghens my suspision. I know thats not what you want to hear, as you have some high dollar PC8500 ram. I would swap test the ram, or try one stick at a time.

 

[InterMurph]

Thanks, GuitarDaddy. I don't have any other RAM to test with it, so I'll try swapping it out one stick at a time.

 

[InterMurph]

Both of my Corsiar XMS208500 sticks failed individually, within 30 seconds each.

And it turns out that I do have some other DDR2 memory; two 512MB sticks rated at PC2-4300. Those each failed within 30 seconds also.

So maybe it's not a memory problem.

 

[aka1nas]

Are you running your RAM at too aggressive timings or too high/low of a voltage?

 

[Roguestar]

Or to test the RAM you could run memtest86.

 

[Duvie]

I hate the myrig thing in anandtech (boy the software on this site sucks)....Tell me what mobo you have?

 

[Roguestar]

Asus P5W DH Deluxe

 

[C Dub]

I'm having the same problem with an E6700 on an Asus P5B. Spontaneous reboots (but I'm running Vista), and it's been up about the same time as yours (4 months). Weirdness.

 

[GuitarDaddy]

If you can verify that the ram is good in another machine, then I would say mem controller on the board is bad

 

[CurseTheSky]

I would agree with memory, PSU, or motherboard related issues first.

However, what kind of cooling at you using on the processor? Is it possible that you used TOO thin a layer of thermal paste, and it all burned up? It may be worth your while to take the heatsink off and check how it looks.

 

[InterMurph]

I finally have some more time to look at this...

First, I triple-checked my BIOS settings, and when I returned everything to their defaults, I accidentally set the DRAM timing to 533 MHz. That should have been OK for the Corsair XMS2 memory, but it clearly wasn't for the slower memory. So I'm retesting the slower memory now.

Second, I will try both sets of memory in another system to see if one is obviously bad.

And third, I will check on the thermal paste, as CurseTheSky recommends.

Thanks!

 

[InterMurph]

Argh! Memtest86+ just ran for 3 hours 51 minutes with the slow RAM and default settings, then it froze up.

Then I hit the reset button, and it started up again. This time it ran for 30 seconds before it froze up.

The next time, it made it 39 seconds before a freeze.

 

[InterMurph]

Then I powered it off for 10 minutes, and started it up again. It lasted 58 seconds this time.

 

[Banzai042]

Could the northbridge be overheating?

 

[InterMurph]

I have a NorthPole fan on it. It's been there for three months or so (running at 365 MHz X 10 = 3.65 GHz), and I can visually verify that it's still spinning.

 

[crispy2010]

Yesterday my pc started doing the same thing. I had cpu and ram oced. Its not a cooling problem, I am watercooled and have fan on northbridge. I put settings back to stock and it stopped I think.!

I thought I just had a virus, or software bug.

What memtest are you using? Is there a version that boots off of cd, I have no floppy!