Didn't want to take over someone elses thread so ill just post my questions here:

How do these idle temps look?

Using a Xigmatek HDT-S1283 with Arctic Silver 5.
The CPU is running at stock speed.



It worries me that one core is considerably hotter than the rest.

Hope someone can advice me what to do.

Thanks in advance.

That's not atypical at all for a Q9450. ...but just for your edification you may want to use RealTemp and check to see if a sensor is stuck. Just for comparison my Q9450's idle temps right now @ a 3.2Ghz are 48, 47, 45, 46

Also, check your vCore voltages. Don't know what kind of motherboard you're using but @ 3.2Ghz I'm stable @ 1.22v...you may need less voltage than what your motherboard auto-selects, which would also reduce temps.

Hey folks, thank you for the quick replies.

I'm using an Asus P5Q-E board. CPU-Z reports a core voltage of 1.144-1.168V. Think SpeedStep is changing this (and the multiplier) back and forth.

This is how Real Temp looks.

My real concern was that i screwed something up seating the CPU as it's the first time i have done this.
You calmed me down, thanks

That idle temperature is horrible for stock q9450 on your high performing cooler at that low vcore.

Run prime95 for at least an hour and check temperature.
http://files.extremeoverclocking.com/file.php?f=103

Dagger, so harsh, its not that bad. However, its not accurate either. The 45nm chips need to be calibrated for more accurate temps. Their sensors are funky. Twattle's real temp is not what likely is shown, but its common for an uncalibrated Q9450. The 45nm temps become more accurate at load, but their idle temps are funky commonly 10-15C lower than shown. Uncalibrated, its not uncommon to see ~50C idles even after the adjustment to TjMax from 105C to 95C.

Twattle if you want more accurate temps you'll need to calibrate. You can use Computronix' guide to do so: http://www.tomshardware.com/forum/ [...] ure-guide.

Keep this in mind too. Your TjMax (the processors highest rated temp) is widely believed to be 95C. If you're core temps are hitting 65C under load that means you still have a 30C delta to TjMax. ...more than safe. For older chips the TjMax value was 75C. ...so if you had 60-65C load temps it was a bit more alarming.

Cheers, i'll read through that guide. I only plan on doing a slight overclock to 3.2 GHz so my only concern is the chip health.
I don't really care about the temperature relative to reviews of that same cooler as long as it keeps me in a safe zone.

Returning when I did the calibration.

Guys don't worry about those idle temps. Not a single program can accurately gauge the temps of the Q9450 without any calibration. If you have Speedstep on and your cpu is running at 6xFSB with no load and you cooler running on high you idle temps should only be about 5-7 degrees higher than your ambient temp. Read this http://www.techpowerup.com/realtemp/docs.php, I calibrated Real Temp to -2 on each core and now its much more reasonable. If anything when you are overclocking you should not worry about the temps of your 45nm cpu unless you are giving it insane amounts of voltage. You will most likely reach an FSB wall (for me around 450) before your cpu starts to overheat. Hope that helps.

yeap- I wouldn't worry about it at all. real temp shows me in that range too- mid 40s peaking at mid 50s while the low budget app that ASUS includes with the mobo always shows around 30 when idle. The Nvidia system monitor is usually somewhere in the middle. Moe's right- there's nothing accurate without calibration but you've got no worries either way if you're not bumping up the voltage much.

Way too high dude.



8.5x333 (with speedstep). CPU Vcc for mine is 1.024 V (real in CPU-Z). Idle temps are the same when running 8.5x400 as well (Vcc there is 1.1250 V).

You might need to lap the HS and CPU.

...what are you load temps?

You're not even running half of the processor that the OP's running though, of course your temps are much cooler.