Acceptable temp for Athlon II

  • Thread starter
  • Admin
  • #1

Tony D

My sister Mary brought her HP desktop to me because it was shutting down. She suspected it was heat-related.

I confirmed her suspicions. About 10 minutes after booting, the CPU fan would spin faster. I installed SpeedFan which said the GPU was running at 98C.

Opened the case and found the CPU heat sink was clogged. Cleaned it out. SpeedFan now said that GPU was running at 72C which I thought was high.

I then installed Speccy and its temp readout is different. See below.

These temps look good to me. Why the conflict between Speccy and SpeedFan? Maybe SpeedFan hasn't kept up with new hardware.

Operating System
Microsoft Windows 7 Home Premium 64-bit SP1
AMD Athlon II X4 630 30 °C
Propus 45nm Technology
6.00GB Dual-Channel DDR3 @ 533MHz (7-7-7-20)
COMPAQ FP5315 (1024x768@60Hz)
256MB GeForce 9100 (HP)
Hard Drives
149GB Seagate ST3160812AS ATA Device (SATA) 33 °C
Optical Drives
hp CDDVDW TS-H653R ATA Device
High Definition Audio Device
Last edited:


Active Member
These sensors are very cheap devices and so I agree with Dougie, the quality of the sensor is huge factor. But I note these sensors simply supply some hexadecimal number to the system and it is up the to monitoring program to convert that number to a temperature value. There is no magic or secret formula, or unique algorithm each program uses. The maker of the sensor should have published specs that say how to interpret the value provided. For that reason, I find it irritating two different programs monitoring the same sensor don't always show the same (or at least close) values.

Some variances make sense because temperatures fluctuations can be extreme very rapidly and these sensors have sample rates that vary. And each of the monitoring programs has different sample rates too, as well as sampling start times. So temperatures shown by Speccy may be of a sample taken several seconds apart from HWMonitor or Speedfan, for examples. The CPU temp can easily vary 10 or 15°C or more in 3 or 4 seconds. That's why I like to monitor my temps in my system tray over time instead of taking a snapshot of it.

Another problem is, there is no industry standard between AMD and Intel, or the board makers saying what sensors will be used, where they will be placed, what their sample rate will be, or even what label will be applied to them. :real_anger:

So it is no wonder there are so many inconsistencies between the various monitoring programs. I like SpeedFan but it is notorious for putting the wrong label to sensors. Speccy is great for system information but as Cindy noted, not so great for "health" information like temps and voltages. We've seen in other threads where Speccy has been WAY OFF in voltages. For example it is showing my +5V at +3.367 and +12VDC is currently at .048VDC. My system would not be running with those voltages so they obviously cannot be right. This is why we, as helpers, must not recommend a user replace a power supply based on Speccy, or any software based monitoring program, alone.

I have found that the voltage and temperature information found in the BIOS Setup Menu (if provided) to be the most accurate but the problem there is running the BIOS Setup Menu is about the least demanding task we can ask of our computers. So that is of no help when we really need to see how our systems are doing when tasked.

I'm with bassfisher and like HWiNFO64 best - especially for voltages. But it provides so much information, it can cause information overload. And it does not provide a System Tray applet. So I use CoreTemp for that.


Anymore most will use speedfan for fan control instead of temps,l I agree with Bill that in most cases the motherboards BIOS will be where you want to go for accurate temperature readings. I go back and fort between CoreTemp and HWMonitor as far as software.
Last edited:


Active Member
I have got so many different readings of temps with software, I have resorted to putting my hand on the heatsink. I can tell if it is running on the hot side.
I haven't the time to keep going into the BIOS.


Active Member
I personally would not trust my hand. A warm heatsink could just mean the heatsink is working very efficiently. Plus the fan, size and shape of the heatsink, specific CPU in use, not to mention ambient temps affect that too much for me. And I might have just been holding my cup of hot coffee, or my glass of ice cold Jack and Diet! But that's just me.

I agree about too much time to go into the BIOS. Plus, as I noted before, running the BIOS Setup Menu is the least demanding task we can ask of our computers. And since a CPU's temp can drop from excessively high to downright chilly in just a couple clock cycles (MUCH faster than it takes to reboot into the BIOS Setup menu), unless the temps already read hot in the BIOS, monitoring temps via the BIOS is pretty useless.

I generally trust CoreTemp and HWiNFO64 for actual temps - and the stability and speed (throttling) of the system make good clues too. I have also used my LASER guided IR Thermometer gun (great also for frying pans, grills, freezers, fridges and dog's noses).

And if really ambitious and I want truly accurate (±1.0%) reading, I have put the thermocouple that came with my Mastech between the CPU and heatsink. This is actually how I found CoreTemp and HWiNFO64 to be most accurate and more importantly, consistent - at least for Intels (I have not used AMDs in any builds since the Core 2 Duo's came out so never checked AMDs).


Active Member
Never had one of my computers shut down from a heat issue, so must be doing something right.
One is O/C to 4.7 Gigs, water cooled and stable.
As cranky Franky sang "I did it my way"
Do what works for you.:)


Active Member
I've had a couple aftermarket cooler fans suddenly seize on me :( over the years with no warning (increased bearing whine) so I cannot say I never had a computer shut down due to heat. But you can tell by the last line in my sig I am a bit anal about heat. I am not a believer that the coolest temps possible are necessary. But maintaining temps comfortably within "normal" operating ranges are absolutely essential. And because CPU temps can skyrocket in just a couple clock cycles, for me, real-time monitoring is essential.

Many sudden shutdowns, reboots, and system freezes (an ironic term, IMO) are due to heat issues. Without real-time monitoring, it is harder to isolate the cause because those skyrocketed temps can and do plummet just as quickly.

We see users complain all the time about "random shutdowns" or "random reboots". I say there is no such thing as "random" shutdowns. There is always a reason and a pattern - though we often cannot see it if we are not looking in the right place.

Water cooling is great is as long as the cooling requirements for the other heat sensitive/heat generating devices surrounding the CPU socket and elsewhere on the motherboard are not neglected - as is, sadly, often the case when alternative cooling solutions are used. An over-heated chipset ("system" temp), RAM, drive, or other sensitive motherboard component can cause a sudden, unexpected shutdown, reboot, or "halt" state (system freeze) too. But unfortunately, many less experienced overclockers and alternative cooling adapters fail to do their homework and fail to properly address case/system cooling too.

"Heat is the bane of all electronics!" That does not imply temperatures have to extend into the "hot" range to cause damage. Running electronics at extremely "warm" temperatures for extended periods of time will increase aging and shorten the life expectancy of the components by making them weaker and sooner, rather than later, less able to tolerate even very "warm" temperatures.

Damage due to heat is often like excessively loud noise and music to our hearing. It may not be loud enough to pop our eardrums and cause instant deafness; but permanent damage to our hearing still occurs, is cumulative, and will degrade our hearing ability faster than with someone who always wears ear protection.