Ok, so I have this problem which I don't quite know what to do about.
By way of background; I have owned this computer for around 3.5 years. In all this time, it's worked flawlessly - perfectly stable, no glitches, everything smooth as butter.
Specs:
I5 705 2.66 OC'd to 3.2
4Gigs DDR3 1333
1 Terra
Micro ATX
GTX 560 Ti
So, middle of the road these days, but with some tweaking perfectly capable of running Crysis / Bioshock etc of good settings.
6 Months after I first got it, on a whim, I installed Coretemp and found it reading a whopping 90 degrees at idle! I thought surely this was a sensor malfunction (I had had it for 6 months on no errors right?), so set it aside and thought nothing of it.
Fast forward 3 years later and for some reason I decide to dl coretemp again.... and I get the same reading. THIS time, I also checked the BIOS AND real temp (why I didn't do this before, I will never know)... and all reads the same.
So clearly, the guy that set this comp up stuffed up the HS/Fan seating...
My question is... WTF do I do?
It's been 3.5 years of this CPU running at this temp and not a raised BIOS alert, no temp monitor alarms... nothing. Computer runs perfectly fine (I just had a 4 hour Bioshock session... no problem).I have even run Prime 95 with this thing for 4 hours testing GFX OC settings with no issues...
Is this even possible? Surely the CPU should have cracked ages ago? I am scared that if I mess with the thing it will come apart completely... could the BIOS Core Temp and Real Temp all be wrong? I mean, its OC'd as well... surely that would have pushed it past breaking point if the numbers are right?
Posts
90C is really high though. I think the absolute max temperature before that chip is dead is 111C, but anything past 80 is danger zone in terms of shortening the life of it.
If your BIOS allows it, save your current voltage and clock settings to a profile, then reset to factory defaults. If this doesn't help then your heatsink/fan are probably fucked somehow.
If you can achieve reasonable idle temps by resetting your system to factory defaults, then you should just work out a stable OC on your own. Shouldn't be too hard since at that point it would mean that it's just overvolted too much.
1. Sensor is broke; OR
2. HSF not seated correctly/at all.
So... I don't know. Maybe its worth just leaving it... worked fine so far.
I would expect the temp to fluctuate if the sensor is working correctly based on CPU load.