Temperature information

I'm hoping some people can help me figure out whether this would be good or not.

Basically I've been messing around with the OCing of my CPU since I've finally got all my temperature issues sorted. Firstly, I notice that my CPU temps at times are erratic, jumping/falling up to 10C at a time. Is this normal?

Secondly, my CPU temps after a short gaming session were
CPU1: 66C/ CPU2: 80C
This seems a bit strange to me. Does anyone know why CPU2 might be so much warmer?

Thirdly, related to the above, is 80C @ 3.4GHz good, bad or average? If I was running at 80C for hours can I expect this to do any damage? I know the CPU can handle up to 105C, however that doesn't mean something below can't cause damage. Confirmation on this would be great

Finally (YAY!) is the airplane strength fans really necessary when idle? Certainly it has its advantages - right now my GPUs are at 47C (the lowest I've EVER seen them) and CPUs are at 38C (Again, lowest I've ever seen them) - however the noise scares me slightly. Is there no way to make it change depending on the reading as it usually would? As nice as a constant 38C is, I see no reason to have the fans on this loud all the time if they could reduce slightly and increase the temps slightly =-P


Thanks for any advice. I think I know the answer to a few of these questions, but I always like confirmation from others =-)

 

Are you using HWMonitor? Is this an aftermarket x9000 (China eBay item)?

I would be a bit worried about the differences between cores. Seems a bit off to me.

If you are using Everest, change the polling period for the sensors (decrease it).

 

HWMonitor, and nope it came from Dell.

 

According to intel, you're fine. Higher temps reduce the lifetime of your CPU though, but I can't tell you how much. Even at 100°C it should at least last as long as your processor warranty lasts (2 years?).

To have HWMonitor report the correct temperature for 45nm CPUs, you have to edit hwmonitorw.ini and change:
CPU_0_TJMAX=10 0.0 -> CPU_0_TJMAX=10 5.0

Otherwise X9000 temps are reported 5°C to low.

I recently switched from a Merom 65nm CPU (T7500) to a Penryn 45nm CPU (T8300) as well. One thing I noticed are that the DTS temperature sensors (those that HWMonitor displays) switch much more rapid and temperature graphs have higher slopes. Instant changes +10°C are not uncommon, while those barely happend with the 65nm CPU before. I can't tell whether this is due to the die shrink that the sensors are now closer to the heat spots or whether intel changed somethin in the implementation of the sensors (e.g. one DTS reading corresponds to an internal reading of many, many sensors and the maximum is reported. One possibility is that intel changed the total number of sensors).

 

7oby, thanks a lot for your insight. =-)

Edit: What about the chipset - what sort of temps should be worrying me? Find attached a screenshot of my current temps @ 3.4GHz. The hottest temps are from when I was playing Prototype - everything maxed out, 2xAA. As you can see my minimum temperatures are fantastic, my GPUs are running cooler than I ever expected they could, even under load, and the CPU is hitting 74C max (Which I don't feel it half bad when overclocked)

The chipset however is currently the warmest temperature at 80C. I've tried looking around but can't find any guidelines for what the min/max temperatures should be on this, and where I'm doing damage. Advice?