Previously stable overclock broken with windows update?

Strawberry · 14 Mar 2010 at 12:39

Hmm..

My previous overclock, at 4 GHz, was completely stable. However last night my PC automatically did a Windows Update. This morning I woke up to a frozen screen, and when rebooting my overclock was completely unstable.

So I tried turning it down to 3.8, 3.6 and 3.4 in succession. Only 3.4 was stable.
So I boot in 3.4 and do a system restore to before the update. Booted back into 3.8, still unstable.

Checked all my temperatures. Northbridge 50°C, Southbridge 30°C, HDD 36°C, GPU 39°C, CPU 37°C.

Did a CMOS clear and re-did all my settings. Worked fine at 3.4, failed at 3.6.
Am running in 3.4 right now, but I would really like my 4.0 GHz back!

Any ideas?

Strawberry · 14 Mar 2010 at 13:42

Forgot to mention - Windows 7 64 bit version. No idea what the update did exactly but it was listed as a "Critical update" or "Security update" or something.

I don't have any spare HDD I can test with atm, and I don't have a DVD drive so I can't install Windows in the first place.

Maybe the windows update messed up my hardware in a way that stayed even after the system restore? Also note it is perfectly stable within my BIOS or running memtest, lunix etc. just windows is completely unstable above 3.4.

Strawberry · 14 Mar 2010 at 14:34

I've passed overnight prime95 tests on 3.4, 3.6, 3.8 and 4.0 GHz, several passes of memtest always worked fine, even if I had a slightly dodgy OC (before I found a stable one), it usually took ~2 hours to error out.

But now the system hangs, crashes, BSODs, programs go crazy, everything fails loading, prime95 errors after several seconds, so does memtest etc. if I go above 3.4 GHz.

3.4 works 100% fine still no error.

Even if the system was slightly unstable to begin with there's a huge difference between one calculation error per 30 hours and hanging crashing erroring after several seconds.

Strawberry · 14 Mar 2010 at 14:48

.walls said:
There are plenty of reasons that you might not be getting the same clocks as you used to (chip degradation due to excess voltages, updated bios and blindly applied the same overclock without retesting, etc, etc). The windows update is coincidental.

Hmm..

It just seems like an oddly big coincidence considering that my logs show the PC crashed just after the update finished.