[qasdfdsaq]

Originally Posted by deakin

Are you thinking of GPU's ?

As the temps you mentioned would scare the bejesus out of me.
I'm running a AMD X6 1035T and even after a hearty game session
i never get any higher than 38-40'c on stock air cooling.
Right now, i'm sitting idle at 15'c in quite a warm room.

Nope. CPUs. GPUs are designed to work at 95'c constantly.

Perhaps you're just paranoid. Intel clearly specify a maximum CPU casing temperature of above 70'c on most processors.

15'c idle in a "warm room" is impossible on air cooling. You would have to be using refrigerated compressant or a peltier to reach that low. A "warm room" is 20-25'c and you cannot get any temperature lower than room temperature with conventional cooling methods.

Advanced datacentres and servers these days are designed around an ambient air temperature of up to 35'c.

The numbers you state bear no resemblance to reality.

---------- Post added at 20:23 ---------- Previous post was at 20:16 ----------

Originally Posted by alanbjames

I really dont think Bios is reading Temp Correctly.

Bios says 62*c
Clickbios in windows says 53*c
Coretemp says 35*c

Just checked MSI control center that lets u overclock in windows and thats only showing temp of CPU at 47*c where Click Bios is 52*c and they are reading from the same board and chip.

Things are not adding up.

I do have some Akasa 455 thermal compound here if it needs changing.

The temperature sensor in the AMD FX processors are known to be inaccurate below 40'c. But then again any temperature below 40'c doesn't even really need to be measured given as far as silicon is concerned it might as well be room temperature.

Note that BIOS and core temperatures are not measuring the same thing. The core will likely vary between being the hottest and the coolest part of the processor depending on what you're doing.

---------- Post added at 20:26 ---------- Previous post was at 20:23 ----------

Originally Posted by damien c

The 8320's are hot chips and it's down to the voltages etc on them, it makes me laugh these days as chips are getting hotter and hotter when they should be getting cooler.

While this is true of AMD chips it's not true of all chips in general. The majority across the market are getting cooler. Chips of 130w or so used to be common in the mainstream segment with anything up to 160w or more in the performance segment.

That's mostly now dropped, a few years ago 95w became mainstream with high-end being around 130w (e.g. 1st/2nd gen i7), most recently last year's Intel architectures reduced mainstream power consumption to 55-65w and 80-90w in the performance segment.

AMD have been behind pretty much the entire decade both in terms of power efficiency and TDP.

---------- Post added at 20:26 ---------- Previous post was at 20:26 ----------

Originally Posted by tizmeinnit

I think you should just stop worrying about it unless it becomes an issue

This.

Any temperature below 60'c is not an issue.

It's impossible to even set the temperature target or alarm below 65'c on my desktop motherboard! The only options are 65, 70, 75, 80, or 85.

On my server, one of the most power efficient (and coolest idle) CPUs available is idling at 39'c. The non-critical high temperature alarm is set at 91'c and cannot be changed.

---------- Post added at 20:38 ---------- Previous post was at 20:26 ----------

Originally Posted by heero_yuy

Most chips are rated to 150C junction temperature (On-Chip) so any reduction will extend life.

If you want to overclock cooler means faster.

Also, this.

(Although I know some non-CPU chips in modern phones are rated at 120'c-140'c)

The maximum temperatures listed in spec sheets and in BIOS alarms are already highly conservative.

The only time reduced temperatures will affect reliability or lifetime is during moderate to extreme overclocking.

Also remember all modern processors have built-in safety mechanisms that throttle or power off the processor if temperatures reach anywhere near dangerous. These are built-in the core hardware and cannot be turned off, and are specifically there to prevent any temperature induced damage. In some cases it'd still be entirely safe to run the CPU with the heatsink removed entirely. Though it'd be pretty damn slow.