E8400 faulty temperatures?

Temperature has been fixed, but please read my last post, I'm having troubles, my processor didn't really like 1.6v I think.

Okay...so yesterday I noticed my one month old E8400 OC'ed @3.6ghz at stock voltage (2.4 at idle, multiplier goes to 6x at idle by itself) is showing some, least to say, remarkable temperatures.
For one the idle temparatures I'm reading out are err...huge, and core #2 is showing 15c higher than core #1, which comes down to 39C for core 1 and 54C for core 2.
It gets fun when I put the thing under load though, temps soaring to ~65 on 1 and ~75 on core 2. This is with ambient temps of around 23C.
My cpu cooler is an Arctic Freezer 7 pro running at ~1500rpm idle and ~2450/2500 under 100% cpu load.
I already made sure the pc isn't reusing it's own hot air and tested with stock clocks, but the high readings stay.
Finally, I took a screeny from cpu-z so you all can check precise stats.

well for one your vcore is way to high for that clock! 2nd reseat your heatsink you didnt put the tim right or something thats why your getting weird temps if that doesnt work lap your heatsink and if that doesnt work lap your cpu... a good way to determin if your cpu/heatsink needs laping is

spread a thin coat of tim on your heatsink/cpu then go to your mirror and press the cpu against it and when you take it off you will see exactly what parts of the cpu are making contact via the tim on your mirror....

and if lapping doesnt work then you have a bad temp sensor

Thanks I honestly have no clue how my voltage got that high, just reset it to default as well as clock speeds and my huge load temps are gone, however, my idle temperatures are still the same. So is the diffirence for the 2 cores, it even got a little larger. I'm suspecting a sensor gone weird there though...

glad u fixed the issue

Wow almost 1.6V, thats CPU degredation territory right there, auto voltage can be deadly with 45nm CPUs.

What are you using to monitor temperatures RealTemp? You've read about the temperature sensor problems correct? What voltage are you using now that you've set it manualy btw.

Edit AGAIN: Re-oc'ed to 3.6ghz, this time on 1.3v, ran some stability and heat checks (still have to run a 3dmark though), put it under 100% for near 8 mins (if anything is seriously wrong it would've shown by then methinks), everything goes as it's supposed to. I still have to check if my processor has good contact with the cooling block, but that's something for later, it doesn't reach over 45c under load (realtemp) so I'm happy, might even be able to get out another 200mhz or something without trouble
P.S: those cpu temps in everest are still 10c too high apparently, but notice the blue one above all others.



Yes I've read about the problem before, so a part of my high temps is probably caused by that.
Using default settings right now, 1.2250v and 3ghz.
However I'm still reading out high idle on core 2. Well, at least it doesn't rise when I load it.

Few things, you know about vdrop and vdroop right? The voltage you set in your BIOs will never actually be recieved by the CPU, first vdrop will be giving it less voltage under idle, and then vdroop will further lower the voltage under load, depending on different motherboards this can be as high as 0.1V or more difference. You can use CPU-Z to monitor your voltage under load, the lowest number it gives you is the least ammount of voltage being recieved under load, and is your effective voltage, or actual voltage.

Secondly, you want to use something good to stress the CPU, such as Orthos, and it definetly needs to be longer then 8 minutes to be deemed 100% stable. Minimum a couple hours.

Also notice how all your screen shots have your multiplier at 6x giving you an effective 2.4ghz clock speed? You want to disable EIST (speed step) so it won't drop your multi during idle, and disable C1E so it doesn't adjust your voltage on the fly.

You shouldn't worry about idle temps as you probably just have a stuck sensor (realtemp has a partialy working stuck sensor test) load temps are really what matters.

Load your OC up with Orthos for an hour and take some screenshots then, all the other ones you provided are kinda useless.

are you sure your heatsink is mounted right... i really doubt one of your temp sensors is stuck what are the chances of that.... if it is rma it!

Damn...autovoltage must've beaten up my CPU pretty bad , as soon as I start using orthos (with overclock), CPU is stressed anywhere from 5-15 seconds, then freezes and reboots the pc (it doesn't do that with everest stressing, but I suppose it's a diffirent stress). I tried with increments of .1v from stock (1.225) to 3.6, always the same results. Except it wouldn't boot at the lowest voltages.
AND one of my sensors is stuck, at least if I believe realtemp, ran the test 15 times and it only scored a 3 in one of them, all others were 0-2, the other one is always somewhere from 4-7, so that isn't too well either I suppose.
Oh yeah, I did disable EIST and C1E.

I am yet to try orthos on stock speeds (I hope it's at least stable there) but I'm not having good hopes, anything you guys can recommend, or am I screwed?