• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Q6600 running too hot?

T

Tudz

Supreme [H]ardness
Joined
Jun 15, 2008
Messages
7,434
G'day,

At work we got a bunch of new 2nd hand computers. They are Q6600s running at stock speeds and I was running simulations on them, after much pain getting them to run multicore via MPI, I thought I'd check out the temps using CPUID HWMonitor.

To my surprise after just a few minutes they were running at 75-80 degrees, which kinda scared me a bit. This is with stock cooling, no overclock at all, and 100% CPU usage running simulations.

Is that too hot for a Q6600? I know my E6600 used to give warnings at 60 degrees, and my i5 maxes out at about 65 degrees (with the stock intel fan screaming like a bitch... really need to upgrade that cooler).

Given we just got these computers I really dont want to fry them!

Cheers,
Tudz
 
That is indeed very hot. At a minimum I would remove the stock coolers and apply some good TIM.
 
What would be a safe temperature for them? At the moment the CPU cooler draws air through a hole in the side of the case. We could mount another fan in that slot, which would blow directly down onto the CPU.

I should probably crack them open and just check they aren't too dusty. The fan seems to be spinning ok, but didn't ramp up very much even when it was over 70 degrees, so maybe there's some quiet computing fan throttling program running?

Because they are work computers I dont want to tamper with them too much. If I open them and then they break at a later date through some unrelated fault I'll most likely be blamed. Probably best just to contact the tech support guys, but historically they like to wait until things are catching on fire before they do anything!
 
No it is not hot and is about right with the stock cooler, 100% load and the EIST cooling vs noise options on in the bios.

The Q6600 has a TjMax (core temp) of around 95-100C, at that temp the CPU will send out an "I am too hot" signal aka PROCHOT# interrupt to the motherboard to spin the cooling fan to 100% and the CPU will begin to reduce voltage and speed to reduce heat, aka throttling. At some temp above TjMax , we do not know for sure as Intel has never said but 10-20C would be a good guess, the CPU will shut itself down to prevent damage.

With Enhanced Intel Speed sTep the bios monitors the CPU temp and adjusts fan speed automatically. In many instances EIST will let the CPU run a bit hotter than most of us like in order to keep fan noise down. 70C at full load still has almost 30C of thermal headroom so the CPU is not even breaking a sweat. What I am trying to say is you must be careful comparing the temps seen on your stock system with temps seen here. Most people OCing disable EIST and run the fans at full speed or use other coolers whose fans are a bit more responsive and/or heatsinks much larger and more efficient.

Your machines are doing well. Adding the other fan in the case side is an excellent idea for increased stability and reliability if during normal usage they are going to be working as hard as your testing. The side fan will like drop temps by 5-10C easy. Go for the added side fan, well worth it for piece of mind.
 
Last edited:
Actually a few FACTS from Intel's site will validate that the CPU's are running too hot.

There are two stepping revisions for the Q6600 processors, a G0 and a B3. The B3 thermal Design is 62.2c as seen here.
The G0 stepping has a thermal design of 71c as seen here.

What is thermal design? Intel explains it best.... "The maximum amount of heat which a thermal solution must be able to dissipate from the processor so that the processor will operate under normal operating conditions." So if you are seeing temps in the 80c range, your thermal solution is not cooling enough to satisfy the thermal design of the processor.
 
You are confusing Tcase with Tjmax. Read the Thermal and Mechanical Guidelines and the data sheet for the 6000 family of processors. Tcase is unusable for us. .

If you go to the page you linked and click on the definition of Thermal Specification you will get this (below). Careful reading will tell you the 70C is Tcase and how and where it is measured. We don't care about Tcase as unless you mill a slot in the IHS and install a thermal probe like the special testing CPU HardOCP uses for heatsink evaluations (the details and method of doing this special thermistor installation is detailed in the Intel Mechanical and Thermal Design Guidelines document ) Tcase cannot be obtained by us and is useless.

Tj (Junction Temp) is what is reported by the DTS sensors embedded directly in the CPU die and is what speedfan/Coretemp/Realtemp/Everest etc report. This reading is easily available to us and is the value the CPU itself uses to determine when throttling, 100% cpu fan speed and catastrophic thermal shutdown should occur.

The following (quoted below) from Intel is a bit misleading as processors with an integrated heat spreader also use Intel Thermal Monitor and report Tj to the motherboard. Unfortunately it is old wording that needs to be updated. It was created when the DTS was first introduced in the mobile version of the P4 which evolved into the Conroe's which were the first C2D's and has not been updated since. The 70C is accurate but only if you can modify your CPU and there is no need to go to the expense and risk as the internal DTS gives us Tj for all modern Intel CPUs.

Thermal Specification: The thermal specification shown is the maximum case temperature at the maximum Thermal Design Power (TDP) value for that processor. It is measured at the geometric center on the topside of the processor integrated heat spreader. For processors without integrated heat spreaders such as mobile processors, the thermal specification is referred to as the junction temperature (Tj). The maximum junction temperature is defined by an activation of the processor Intel® Thermal Monitor. The Intel Thermal Monitor’s automatic mode is used to indicate that the maximum TJ has been reached.
Click to expand...

More background.
http://hardforum.com/showthread.php?p=1031827510#post1031827510

Post #31 is the meat.

Also you might read the program notes/description of Realtemp and Coretemp which describe where and how the temp numbers reported are obtained.

More.
http://hardforum.com/showthread.php?p=1031080147#post1031080147
(BTW the Unclewebb posting in that thread is the author of Realtemp and the only guy I know who had balls enough to run his CPU without a heatsink and on purpose make it throttle and then shutdown so he could empirically determine via the DTS what TjMax and TjCatastropic where for his CPU. )

Intel has never published the Tjmax for the 65nm cores. We determined them by testings and some back door comments by Intel.

At the 2008 IDF there was a presentation and values of Tjmax (unfortunately the presentation was taken off of Intel's site) for the 45nm CPUs. I have a copy but no way to post it and its Intel property anyway. The values of TjMax for the 45's as reported by Intel are:
http://www.anandtech.com/weblog/showpost.aspx?i=478

We went through all this in 2007. The "facts" are here:
http://www.intel.com/design/core2quad/documentation.htm
 
Last edited:
You must log in or register to reply here.
Back
Top