My CPU is 60C at idle - what do I do?

[limitup]

Hmm. I've been mainly using Core Temp to monitor the temps, so I did some Googling. I read that in some situations Core Temp reports temps that are up to 10 degrees higher than they really are, so I started looking at some other utils.

A lot of people seem to like SpeedFan and some call it the most accurate, so I tried that next. It reports the exact same temps for all 4 cores.

(My computer does NOT like SpeedFan. It basically locked up my computer for about 60+ seconds when I first loaded it, to the point where I was ready to force a shutdown. Then all of a sudden it started working. Or maybe this is normal ... ?)

Fan Control, which is a program made specifically for Dells, also reports the exact same temps.

Everest also reports the exact same temps.

Finally I tried Real Temp 2.60 and now we have a real discrepancy. Real Temp shows temps that are roughly 15C less than all the other utils.

I get no throttling when my temps are over 71C (according to all the utils aside from Real Temp), so do I assume that Real Temp is correct and all the others are wrong?

I ran another stress test at 100% load for about 15 mins and the highest it got according to Real Temp was 60C while the others all show 75C as you can see in this screenshot. At this point the fans were not even on high either.

I did the "Test Sensors" function in Real Temp and it indicates they are all good, and according to Real Temp my idle temp is about 34C.

 

[limitup]

Ok well upon further investigation I found out that Real Temp is using a TJ Max of 85 for my CPU, and the others are all apparentely using a TJ Max of 100, which explains the 15 degree difference.

As a noob I obviously have no idea which is correct, but from the quick research I've done my QX6700 B3 stepping CPU should have a TJ Max of 100, not 85.

To confuse things even more, Dagger said I should see throttling at 72C. But from what I know throttling happens 5 degrees before TJ Max, so whether my TJ Max is 85 or 100 that still doesn't make sense to me...

On the plus side, with my old heatsink the fans would go nuts after a few minutes at full load. Now even after 30 mins of 100% load they still don't go into high ... so I assume the CPU isn't getting TOO hot and I should be safe.

 

[halcyon]

I rather doubt that 100C is the accurate TjMax. 85C is much more believable to me.

 

[dagger]

I rather doubt that 100C is the accurate TjMax. 85C is much more believable to me.

That's wishful thinking. It should be more like 90 or even 95. Besides, it's better to overestimate, just to be on the safe side. Coretemp is what I always use for q6600, and it registers 58C full load prime95 small ffts. You want it cool... 😛

 

[halcyon]

Well, the newer 45nm chips are thought to have a TjMax of 95...so I'd think the older processor would have a lower TjMax as that seems consistent with older processor. In fact his QX6700 shouldn't have a TjMax too different from yours Dagger...just seems more logical that it'd be 80...85C. What's Core Temp estimate your TjMax at Dagger? Isn't the QX6700 in the same generation as the Q6600?

 

[dagger]

Well, the newer 45nm chips are thought to have a TjMax of 95...so I'd think the older processor would have a lower TjMax as that seems consistent with older processor. In fact his QX6700 shouldn't have a TjMax too different from yours Dagger...just seems more logical that it'd be 80...85C. What's Core Temp estimate your TjMax at Dagger? Isn't the QX6700 in the same generation as the Q6600?

It's using 100C. 😛

Strange, it's reading 38-40 on idle across the cores now, compared to 35-37 idle I got 3 month back, when I stressed it. Maybe somehow it's using a lower TjMax then? Oh well. Doesn't matter.

 

[limitup]

Well now I'm pretty confused lol. At the very least, is the fact that I don't see any throttling even when at 100% load for 15+ mins a good sign that I'm not running too hot?

 

[dagger]

Well I'm totally confused now lol. At the very least, is the fact that I don't see any throttling even when at 100% load for 15+ mins a good sign that I'm not running too hot?

Most people run it for 12hrs without errors to ensure stability. 😛

 

[limitup]

What makes no sense is that I'm now running what is basically Dell's most efficient/best heatsink (based on my research). They shipped it with a crappy all aluminum one which I think was a mistake as that heatsink is not made for quad core setups I don't think, so what I have now is how I imagine it should have came from Dell. Yeah my graphics card puts off some heat but the heatsink sitting on top of it is right in the path of the air from the fans and should flow right out the back so that shouldn't make much difference. Maybe Dell's just run hot I dunno. But installing a non-Dell heatsink is a total hack job that not many people attempt, so I have to think my setup should be OK otherwise Dell would figure out a better system, no? I dunno, maybe I can look into a better fan ...

 

[halcyon]

It's using 100C. 😛

Strange, it's reading 38-40 on idle across the cores now, compared to 35-37 idle I got 3 month back, when I stressed it. Maybe somehow it's using a lower TjMax then? Oh well. Doesn't matter.

Dagger, check that out, I think 100C TjMax for a Q6600 sounds a little fishy. Do you have Everest, can you confirm?

 

[limitup]

What do you all consider to be the most accurate util for determining temps? And do you all actually know your accurate TJ Max value etc and set this in the software you use? Or do you just assume whatever util you're using is correct? Maybe all my worrying is for nothing at this point. At least I'm running 10-15C cooler than I was before...

 

[halcyon]

I tend to use 3 utilities, and they're all consistent for me. I use RealTemp, CoreTemp, and Everest Ultimate. Intel does not release TjMax values for its processors to the programmers of the utilities estimate what they believe it to be.