What's your room temp?
It does have an effect on the CPU temp.
It's still a bit wintery over here.
Room temps are 24deg and my CPU 0 idles at 27deg and 41deg at full load according to core temp and the other program I use when at pretty much stock temps.
My CPU 1 temp is always about 2-3deg less so I don't count it.
If the room temp is at 35deg with a fan heater going it will reach 52deg under load no probs.
This was with as5 thermal compund applied.
The max temp used to be a bit higher with the run of the mill compound.
So as5 does reduce max temps.
Yeah that 45deg max for an X2 thing in the OC guide is worrying for the new overclocker.
But realistcally your'e only going to get those max temps on water cooling with an X2.
I am not sure where the 45degC number came from in the OC guide. Seems to me that is much too conservative. I have been unable to get a clear answer on whether the temps reported in BIOS (and therfore in Speedfan, Everest, etc...) are meant to be a silicon temperature or a case temperature. Since the silicon runs hotter than the case then the worst scenario would be if it was a case temperature. It seems to be difficult to find a clear answer on what the maximum allowed die (junction) temperature is for the X2 processors (if anyone can point me to something it would be appreciated). However, for the old Athlon XP processors Tjunction max is 85or 90 degC. Assuming that the X2 are not radically different than this then it seems to me that you could use about the same number for the Tjunction max.
For the uPGA package I would assume about 0.1degC/W junction to case as a reasonable target. Assuming a TDP for the X2 of about 80W then the delta T from case to junction would be 80W X 0.1degC/W = 8degC.
So if you were using 45degC as your absolute max target, and this was actually a case temperature (worst case) then your silicon would be at about 53degC which is about 32degC margin compared to 85degC max. Seems a tad on the conservative side to me.
Well what if you are overclocking? In that case the TDP is going to be higher since we are operating beyond the AMD "P State" as they call it. So let's say you are running at 120W instead of 80W. Then delta T is 12degC so you lose 4 degC margin. No big whoop.
So with my overclocked X2 3800 @ 2.7GHz if I were to assume that I was at TDP of 120W (pretty pessimistic I think but lets assume that for arguments sake). If I am seeing 55degC and that is a case temp then my silicon junction temperature is 55degC + (0.1degC/W X 120W) = 67degC. This is still about 18degC of margin.
So call me crazy but I don't see the big deal of running in the 50's. Spending $90 on a new heatsink or $250 on a water cooling setup to get this down to the low 40's seems to be a waste of money IMHO. Maybe I got something wrong in my calculations or my assumptions so please correct me if I'm wrong.
Cheers