CPU temp, core temp and intel thermal specs...

[gsrcrxsi]

ok so IMO there seems to be some confusion on what people consider "safe" temperatures, and instead of running through threads, i feel it deserves its own thread. many people seem to be following the intel thermal specs for temperature and then applying that to core temperature. but from what i see the thermal spec is for the CPU temp, not core.

It is measured at the geometric center on the topside of the processor integrated heat spreader.

am i the only one who notices this? or am i just crazy?

 

[gsrcrxsi]

thoughts?

 

[prime81]

The core temps directly effect the CPU temp, if the CPU throttles it's due to the core temp sensors. So with that in mind I would say the CPU temp itself is not as important as the cores. Just my thoughts....

 

[jason4207]

You are correct, but I've gotten tired of telling people that the 72*C spec is not the same temp you see in CoreTemp/RealTemp.

You should be able to run these CPU's right up to Tjmax w/o issue, but you will shorten their lifespan if you run them that hot (how much so is debatable). Running the CPU that hot will also limit your OCability. It's hard to find stability at 80*C+ unless you're at or near stock clocks.

72*C is the max IHS temp, but most people tend to agree that 20-30*C to Tjmax is as high as you want to go. Since Tjmax is ~95*C (according to RealTemp), the 72*C temp also lies right in the range that most people deem acceptable.

Transistors operate more efficiently, and can switch faster at lower temps. It's always a good idea to get temps as low as possible.

 

[diaz]

Look at stickies... I made one in "intel" section.

-D