E8400 Temp variable from one chip to another?

[Firstmate]

My question is, are there differences from chip to chip in the native heat generation of the same model CPU? I assembleld my System in the spring of 2007 using an BTF-90 HSF which didn't perform to the specs shown in the Frosty Tech review, but it was within 5 deg.C across the board idle/load. I decided this May to try a different HSF to see if I could cool it down a bit. I'm currently using the Xigmatic Red Scorpion and AS5. This cooled it down about 1 or 2 deg. from the BTF90 but still not up to the listed spec on Frosty tech review for this unit. I'm currently running 45/47C idle and as high as 61/64C when gaming, now with the case side off and a fan blowing in to it. No OC. Ambient 28-30C by infrared sensor aimed at case surface.

 

[GhislainG]

Not all E8400 CPUs are equal; they can vary from one unit to another and even between cores. Does your room temperature match the room temperature at Frosty Tech? If your room is warmer, then you can't expect to get the same results.

 

[randomizer]

The temperature sensors are calibrated, they are not perfect. Each sensor is different so at the exact same actual core temperature you could have many different readings across different CPUs and even the individual cores in the same CPU.

 

[Firstmate]

Thanks for the reply.

It do understanding that the temperature rise displayed in the FrostyTech reviews is above their ambient. My ambient is higher than theirs [about 3 or 4c. ] . My concern is that the rise above my ambient is greater than their tests incidate by 4 or 5 degrees. Their results show about 11 deg rise at idle, above their ambient temp, mine are about 15 to18 deg. above my ambient. This is with the case cover off and a fan blowing room air into the case. I'm using "RealTemp" and the results quoted are T-junction. I find it diffficult to belive that mounting and thermal grease application are the problem. I have attenpted to be very meticulous in cleaning and purifying the mounting surfaces, applying the requisit minimum of AS-5 paste to both surfaces and pressing the push pins into the MB.

This is my second HSF. The first was about the same, with a screw a down mount, thus the question about variation of operational temp. from chip to chip.

By the way I have reviewed the second repliers link on more than one ocassion during the last 2 years and although my load temps remain just in the safe area for Core 2 Duo (--60--/--65--65--Safe], I'm still concerned, since my ability to overclock might be severely limited if I should ever decide to do so.

Any other thoughts would be appreciated.

 

[GhislainG]

Mine idles at 42°C in a room at 22°C and in a closed case. This is approximately 5°C higher than the two Q6600 and the E6420, but Tj Max (105°C) also is 5°C higher. While running Prime95, it reaches 69°C which still is 36°C below Tj Max. If you're worried, install the best CPU cooler that you can find.

 

[Firstmate]

Thanks again, I see your idle is temp 2deg C higher than mine, but with a closed case, similar to mine if were to close my case. This is somewaht reassuring. It appears that the reviewers/testers are somewat removed from the real world when they publish their lab findings on these subjects. I'm currently using one of the highest rated HSF's. The Xigmatic S1283.

 

[GhislainG]

They often test on a bench in a cool room. What's important in a test is how well a cooler performs when compared to the OEM and other coolers.

 

[randomizer]

How do you know your rise above ambient at idle is greater than theirs? You put too much faith into sensors that were never designed for you to check temperatures. You are not using the same CPU so your results are not directly comparable, especially at idle (the sensors lose accuracy the cooler the CPU gets).