Intel Core i7-930 D0 step at 4.0 GHz
Masscool Long Life CPU cooler w/92 mm fan (not best for OC but worked)
Corsair 6 GB RAM 9-9-9-24 at 1532 MHz (slight underclock to run CPU at 4.0 GHz)
2x Radeon HD 5850 in crossfire
Rosewill 850 W PSU (not the best, but has been OK so far)
WD Caviar Blue 640 GB
Azza Solano Full Tower
I've had my CPU overclocked to 4.0 GHz since at least late Feb. 2011. I would say it was quite stable: I could run Prime95 for 36+ hrs without crashing, 20 runs of max RAM Intel Burn Test without issue, max temps of 78-80 C, but while gaming I never go above 63 C.
For months I've been able to turn my system on, it would fire up without issue, perform like mad, then shut down without issue. Yesterday in the afternoon I shut my system down, came back an hour later and tried to turn it back on but it was in OC recovery mode. I tried raising core voltage, lowering bclock, but the only thing that worked was hitting the 'reset' button which returned the BIOS to stock settings. Fearing my CPU may have been damaged somehow, I ran 10 passes of Intel Burn Test, which took time but was OK (I know it's not Prime95, but I didn't want to wait 24 hrs).
I also took this as a chance to crack open my case and clean out the dust that had collected; nothing major but enough to see some dust here and there.
Unfortunately I did not write anything down from my last OC, so I took a few hours to get back to 4.0 GHz with enough volts to post and run IBT but low enough volts to minimize temperature.
Right now my temps are more or less the same as before, and I can update my volts later if needed on this thread. I didn't want to run Prime95 when away from the house in case it crashes, but it did run 10 passes of Intel Burn Test without issue.
I did search www.google.com and have a few things to test, but I wanted to see what the people at Tom's had to say.
My first thought is that just maybe enough dust had collected to interupt some airflow, but I always monitor CPU temps and they were no higher yesterday than before.
My second thought is the recent bout of bad weather in the US SE region may have affected my surge protector, which in turn may have let a surge through last night. Thankfully from previous experiences I did have the computer unplugged during the storms.
My third thought runs with my second and is that my PSU has been damaged and isn't supplying a constant voltage anymore. I don't have a voltmeter, and honestly I don't have the expertise to use one appropriately but can learn.
Any other possibilities?
I'm willing to spend some money to safeguard my system but am on a tight budget. At most I'm thinking a new UPS and a new PSU may be in order.
In order to really utilize my system at 4.0 GHz for more strenuous activity than gaming I will want to get a Noctua cooler or something along that line down the road, but right now I'm just about making it work.
I'm honestly not even too concerned about maintaining my OC but rather I view the OC as the 'yellow canary' with my system; I feel that if I can't continue with the OC, then something is wrong with my system that needs to be fixed.
Run prime 95 if it's stable then's it good. Check all your temps for the next couple days. Another question does your bios allow you to save an oc do the bios. Cuz that would suck to have to do this over again!
It worked OK for several days, but now about once a day after a cold boot (system fully shut down) it enters OC recovery mode. I then either raise Vcore or drop the bclk down a few notches and everything boost fine and runs 20 passes Intel Burn Test no prob. To make things interesting, I restart, set the bclk back to what failed the first time or drop the voltage back down and presto, the system boots without issue.
If anyone has an idea of how this is possible, or what to check next, please let me know.