**Official e8400/e8500 Retail OC Thread

[unclewebb]

... but from experience it doesn't look right either sometimes the cores are only 6 to 10 degrees hotter under load than the socket temps, and thats for sure not right.

Most socket temps on most motherboards are far from 100% accurate, 100% of the time. That's the whole reason that Intel went to on chip temperature sensors. You also can't compare the difference between socket and core temps when you are using two different processor families like 45nm and 65nm. The 45nm chips are much more compact and could be transferring more heat directly to your socket sensor so the smaller delta might be caused by that. You would need to do some more testing to find out for sure.

Something is wrong with real temps as well.

Just a minute here. Now you're getting personal!

I wrote RealTemp because all of the other temperature monitoring software was out to lunch for my E8400. When RealTemp is properly calibrated, it can track the core temperature of my E8400 within a degree or two from 10C to 100C. It's hard to get any more accurate than that when the DTS sensors were never designed to give out accurate idle data and when the data coming out is only integer values.

Here's a good comparison of temperature software when using my E8400:
http://www.xtremesystems.org/forums/...&postcount=573

I'm presently running a Q6600 and have noticed that one of the four sensors is reporting about 4C lower at idle so I'm planning to add individual calibration for each core. I wasn't planning to include that option for the dual cores since they typically track each other pretty closely but after reading some posts here it looks like I'll be adding that feature to the dual cores as well. The correct TjMax combined with this feature can really help improve the accuracy of your reported core temperatures.