Instant reboot at high temps

[LazyGit]

I've OCed my e6600 to 3.6 GHz (9x401). Speedfan is telling me that in my relatively cool living room my HDD is at 27C, mobo at 35 and CPU at 37.

If I were to run intel's TAT and give the cores a full workload the temps would quickly hit the 60s, this is with a Zalman 9700. The fan speed has almost no impact, whether it is at minimum or maximum it will still only reach the same temps. One core (the left in TAT) is consistently 2-3 degrees higher than the right. After running for a couple of minutes and the temps settling at about 63 degrees the system reboots.

Is my cooler doing anything? The general invariance in temperature with fan speed seems strange. Also, if the temps are not even reaching 65 should it be shutting down? If I run dual sp2004 the same thing happens but there are no failures. Is there some failsafe shutting it down early that I don't know about?
cheers

 

[moosey]

maybe it's just not a stable o/c

 

[drum]

Originally posted by: moosey
maybe it's just not a stable o/c

i'd try backing down that OC and see what happens

 

[LazyGit]

I thought non stable OCs would fail checks like sp2004. drum, if I back down the OC what am I supposed to look for? This rig will run games for hours with no problems, it's only when the CPU is put under a full load and it hits the 60s that it reboots.
cheers

 

[pkrush]

Intel's tool causes the CPU to draw significantly more power than any real-world load (or even CPUBurn, for that matter)will draw. The difference is at least 30 watts. This could easily be enough to cause your chip to overheat and crash.

 

[LazyGit]

I uderstand that if it were to overheat it would crash but I've seen tests on the net where e6600s have hit 70 in testing and they never said they failed. Isn't ~60 degrees a bit low for the CPU to be crashing. Also, dual sp2004 causes the same problem.
cheers

Edit: Okay, sp2004 failed for the first time. I've upped the voltage so that cpu-z is now reading 1.33 and with dual sp2004 it's not failed in the last 12 minutes anyway and the temps have only reached 61.

Thanks guys.
cheers

 

[Roguestar]

12 minutes is nothing on Orthos. Let it run for a few hours at least. One core tends to be slightly hotter than the other because it's running the system idle processes and background OS tasks, nothing unusual.

 

[LazyGit]

Cheers, Roguestar. Orthos ran for about an hour before I had to go so it looks okay so far but I'll push it further. You're running a zalman 9500, what sort of temps do you get (ambient versus core) on that OC?

 

[Blain]

Heat isn't the only cause of instability. Components simply have a limit.
Face it, you're stuck with inferior parts... Time to upgrade AGAIN! :laugh: