Presumably my very pricey QX9650 that died last week did so because of higher voltage supplied--it was 3y 4m in service when it died. I'm doubtful it's heat related because I had core temps running below 60C almost always, and usually in the low to mid 50's. I also use a cooling soln that some have felt might be implicated in some way, which I can imagine but kind of doubt. Mostly it was just an open back of my case with air blowing rather gently on the mainboard, memory and CPU, but at time I would turn the A/C compressor on and get cool/cold air flowing over parts. There was conjecture about rapid temp changes and also static and also microcondensation, but I would maintain those are remote when you look at the temps involved. Plus, A/C is a dehumidifying event, plus we live in a very dry climate.In any case, stock vCore is 1.25, with a stated range of 1.1 to 1.30v I believe. I ran my processor at 1.3625 to maintain a stable machine at 3.90Ghz. For non-FSX logons, I'd often run at default CPU/memory settings, ie 1.25v in the BIOS.The replacement CPU I bought is a Q9650. I'm having super results loosening the memory timings and aiming for higher DRAM frequency, so I'm at 1661Mhz there, and 3.76Ghz clock for the CPU. It appears I need 1.35v to be stable at this o/c. Everest sensors show a CPU voltage of 3.2 when I have 1.35v set in the BIOS, for what that's worth. I can go higher w/ both CPU and DRAM clock frequency, but then it's up with the voltage.Testimony on sustained overclocks in terms of sustained reliability? Quite frankly I was surprised to see the failure I did at this voltage for that period of time, though Intel makes it pretty clear you can't expect sustained reliability when you exceed the 'functional operational limits' (not the MAX limits, that's even worse!). If I get another 3y out of this, that's cool because it will be big upgrade time then.Thanks in advance,Noel

Hi,With good cooling you should be able to go to 1.4V and still be safe.Watch your FSX though because it doesn't take to well to some overclocking, as shown in a very new thread here.

I have a Q9550 that's been running 4GHz @ about 1.35V since August '09, and an E8400 that's also been @ 4GHz at between 1.4-1.45V since Feb '08. Both still going strong.Voltage is unlikely to be your problem, you're well within maximum operating voltage for ambient + temps on a 45nm CPU. Probably your cooling method as others have speculated. Stop doing that and just get some proper case and CPU cooling.

I have a Q9550 that's been running 4GHz @ about 1.35V since August '09, and an E8400 that's also been @ 4GHz at between 1.4-1.45V since Feb '08. Both still going strong.Voltage is unlikely to be your problem, you're well within maximum operating voltage for ambient + temps on a 45nm CPU. Probably your cooling method as others have speculated. Stop doing that and just get some proper case and CPU cooling.

Thanks. Let me ask you a question on what you believe constitutes 'proper' case and CPU cooling. I initially used A/C during the summers, and then only when o/c to 4.1 or higher Ghz. I stopped this practice after a year or so because I simply didn't need the A/C cycling on and off. And w/ o/c's at 3.9Ghz, the less aggressive option I used (no A/C) was adequate to maintain temps below 60C. Having the case cover open, and blowing room air at components. I think it's relevant to note this is high volume, low flow rate air flow which should create no more problems than the very air cooling solutions retail CPUs come with. I could see how maybe cycling on/off A/C could possibly create too rapid a change in temps, but this doesn't happen with the just air flow and no A/C. I'm using the stock retail HSF for QX processors.Is it safe to say 'proper' meets these criteria:1. Keeps CPU well within max temps at all times. As I say, <60C is well within. Temps are measured w/ Core Temp I believe the app is, and jives w/ Everest's sensors as well. Usually I'm at around 56C currently when O/C'd.2. Doesn't generate moisture or static. I don't see how anything I'm doing currently will generate either other these as there is no moisture involved and air flow rate is <= what the retail HSF does.

Your prior cooling solution introduced a flow of charged particles into the system at the very least, debris and condensation at the worst. It is very likely the cause of your CPU failure, considering your relatively low voltage, temperature, and clockspeed. Now if you had said you were running 1.6V or some similarly high number then I would cite that as the likely cause.

Air from your CPU fan introduces charged particles into the system.Air from your CPU fan is not filtered in any way and contains debris.Air conditioned air is actually less humid then non-conditioned air and a few degrees difference in the ambient air temperature is not going to cause condensation at the CPU/MB interface. If this was true then all of us that run A/C in our summer times would be blowing out our CPUs every year.

Noel you are not a newbie here.Please comply with this forum rules.http://forum.avsim.net/topic/329986-hardware-related-threads/