No wonder the system temperature seems high!
- Thread starter Aunty Em
- Start date
One thing I've found about System temps, and CPU temps, don't rely on people that tell you it should be 30 or 34 or whatever, rather, watch your system, and watch for changes. Now that you know what "Normal" is for your system, just watch to see that it doesn't go up alot from there.
Aunty Em
Well-Known Member
Yeah, I've looked at the positioning of all three sensors on this board and the sensors on KT7A and apart from the sensor beneath the cpu they are quite different.
When I compare those I find that this temp is lower than the one on the KT7A, 39C as opposed to 44C, but that doesn't mean that the 1 GHz TBird on that board will have a higher core temp than this one because I believe they don't run as hot as the Thoroughbreds. What it does say to me though is that when I overclock the TBird I probably need to install a temp probe to get a more accurate idea of what's going on.
As for this board, although the temperature may seem high @ 42C it's actually remarkably stable under load. The core temperature itself has come down by 5C since changing to the coolermaster copper heatpipe with the silent 32 cfm led fan and hopefully when the AS3 kicks in will reduce further.
Actually the difference in noise between the two systems is VERY noticable. Power is supplied by a Thermaltake 420W Dual Fan Silent PSU. The only fan I can really hear is the 92mm YS-Tech fan in the front @ 27 dBa - the top fan and rear fan are YS-tech 80mm 20dBa fans and the coolermaster led fan on the heat sink is < 25dBa. I could probably get the processor cooler with a faster fan, but at the cost of the noise, which I can't stand. This one hums pleasantly in the background.
In fact I'm changing the case fans in the other computer to YS-tech 80mm 20 dBa fans as soon as possible. And will use the GlobalWin heatsink when I overclock - 27dBa.
When I compare those I find that this temp is lower than the one on the KT7A, 39C as opposed to 44C, but that doesn't mean that the 1 GHz TBird on that board will have a higher core temp than this one because I believe they don't run as hot as the Thoroughbreds. What it does say to me though is that when I overclock the TBird I probably need to install a temp probe to get a more accurate idea of what's going on.
As for this board, although the temperature may seem high @ 42C it's actually remarkably stable under load. The core temperature itself has come down by 5C since changing to the coolermaster copper heatpipe with the silent 32 cfm led fan and hopefully when the AS3 kicks in will reduce further.
Actually the difference in noise between the two systems is VERY noticable. Power is supplied by a Thermaltake 420W Dual Fan Silent PSU. The only fan I can really hear is the 92mm YS-Tech fan in the front @ 27 dBa - the top fan and rear fan are YS-tech 80mm 20dBa fans and the coolermaster led fan on the heat sink is < 25dBa. I could probably get the processor cooler with a faster fan, but at the cost of the noise, which I can't stand. This one hums pleasantly in the background.
In fact I'm changing the case fans in the other computer to YS-tech 80mm 20 dBa fans as soon as possible. And will use the GlobalWin heatsink when I overclock - 27dBa.
Justintime
Guest
The 1Ghz Athlon dissapiates about 49W at default Vcore, the XP 2200+ (which you have i think?) is about 54W, not 100% sure as my memory ain't that good
Aunty Em
Well-Known Member
So when the literature says that the core temp of the TBird is about 6C higher than the temp recorded from under the processor that's probably about right?
I presume that taking the reading directly from the diode is a fairly new phenomenon and they don't really have enough data to make a generalisation about that at this time, but the difference in mine is about 13 degrees - beneath core 39C and core temp from diode 52C. I don't know if the AS3 will make a difference to that as it takes a minimum 72 hours running time to start working fully. I can't say I'm particularly impressed by it at this point, but maybe I'm just too impatient.
Before I installed this heatsink the sensor under the cpu and the system sensor were recording the same temperature (+/- 0.5)- Now they are 3 - 4C apart. The system temp is about the same as it was before. I presume this is because the heat is being dissipated into the heatsink more efficiently and therefore away from the sensor.
I presume that taking the reading directly from the diode is a fairly new phenomenon and they don't really have enough data to make a generalisation about that at this time, but the difference in mine is about 13 degrees - beneath core 39C and core temp from diode 52C. I don't know if the AS3 will make a difference to that as it takes a minimum 72 hours running time to start working fully. I can't say I'm particularly impressed by it at this point, but maybe I'm just too impatient.
Before I installed this heatsink the sensor under the cpu and the system sensor were recording the same temperature (+/- 0.5)- Now they are 3 - 4C apart. The system temp is about the same as it was before. I presume this is because the heat is being dissipated into the heatsink more efficiently and therefore away from the sensor.
Justintime
Guest
The Tbird has no diode, so the true core temp is not known, it has to rely on the onboard diode, which can never be 100% accurate, although mobo manufacturers use compensation in their calibration, which is why sometimes one BIOS upgrade can vary in temp readings vastly from another, as they may have recalibrated the temperature compensation.
Aunty Em
Well-Known Member
Well they've been sitting at:
system 40C
under cpu 37C
cpu core 50C
Since this morning with an occassional blip when I've been busy, of no more than 1 - 2 C. So I'd say that's fairly stable. I expect them to rise a couple of C this evening when the central heating comes on.
I didn't notice any difference in cpu temps when I upgraded the Bios on my KT7A. You know it seems a bit strange to me that AMD have waited so long to put a temp diode in their processor. Surely the maximum temps on their website are based on core temps, but I've noticed in this bios that the cpu overheat warning is based on the temp of the diode beneath the cpu and not the core temp. ?(
So if that rises above 60C I'll get a warning.
system 40C
under cpu 37C
cpu core 50C
Since this morning with an occassional blip when I've been busy, of no more than 1 - 2 C. So I'd say that's fairly stable. I expect them to rise a couple of C this evening when the central heating comes on.
I didn't notice any difference in cpu temps when I upgraded the Bios on my KT7A. You know it seems a bit strange to me that AMD have waited so long to put a temp diode in their processor. Surely the maximum temps on their website are based on core temps, but I've noticed in this bios that the cpu overheat warning is based on the temp of the diode beneath the cpu and not the core temp. ?(
So if that rises above 60C I'll get a warning.
Justintime
Guest
The temp diode came out since the Palomino XP processor, but you're right, they really shoulda implemented since the debut of the socket or even the SLOT A Athlon.