E8400 Core Temps

[mormonsniper]

Hi everyone,
Just installed and reinstalled my CPU. Core 0 is at 49C and Core 1 is at 36C (HWMonitor64). Am I doing something wrong? I am using an Ultima 90 w/fan. Orientation of the case make no difference (heatsink strain from gravity?) No lapping... yet.

eta: Rosetta at 100% I get 57C and 43C.

TIA,
ms

 

[Wrathre]

I'm idling myself @ about 49-50c myself for both cores. I just set up my system as well, going to reapp the compound/heatsink in a bit and see if that changes, will post my results.

 

[imposter]

mormonsniper,
Sometimes cores are not calibrated correctly, thus strange results (more of an AMD problem)

but a certain batch of E8XX Are known for strange for Coretemp Problems. As of now it is an intel problem and coretemp can do nothing about it. I guess you were in one of those unlucky people. All i can say is download Coretemp the newest coretemp located here. http://shahsmerdis.com/CoreTemp/CoreTemp.zip

Imposter

 

[mormonsniper]

My E8400 is an OEM from Tankguys. I am afraid that this might limit any over clocking I might want to try. TIM is Shin-Etsu X23-7762 Thermal Compound.
I used Articlean to prepare the surfaces of the heat sink and CPU. It appeared the first time I had plenty of coverage without excess thermal compound.

Also, voltages are very strange. 5.12V on the +12V line and so on....

Mobo is a Gigabyle EX38-DS4. 8 GB G.SKill RAM.

Thanks imposter - CoreTemp was 56C and 43C.

 

[gigabit]

Hi everyone,
Just installed and reinstalled my CPU. Core 0 is at 49C and Core 1 is at 36C (HWMonitor64). Am I doing something wrong? I am using an Ultima 90 w/fan. Orientation of the case make no difference (heatsink strain from gravity?) No lapping... yet.

eta: Rosetta at 100% I get 57C and 43C.

TIA,
ms

Im having the same problem with the same setup.Ultima 90.My Ultra 120 is on it way from the heatsinkfactory.com atm.I hope it helps

 

[mormonsniper]

The 120 would not fit in my P182 case, that not withstanding, the difference in temps between the two cores is not related to a bigger heatsink me thinks.

PSU is the PC Power & Cooling 610 by the way.

 

[gigabit]

It could be because one side is sitting on the ihs better

 

[muddocktor]

I've read here that some of the E8400's have a problem with the temp sensor diodes on the chip getting "stuck", but that once your temps get past that point they start reading. But since I don't have an E8400 I can't confirm this. I do know that my 45nm Yorkfield has no problems such as some of you guys are getting with the E8400 procs. But it might have other problems with Core Temp setting the Tj Max too high. There is about a 10 C difference between Core Temp and Real Temp, which corresponds to the Tj Max assumptions of the 2 programs.

 

[OldSkool]

Mine idles around 32c both cores and is around 45-52c full load. I'm using a TRUE120 and although that will lead to lower temps all around, I doubt it evens out the cores. I think if you just reapply the tim and try to reseat the hs it may help. Try a different orientation of the hs if you can.

 

[mormonsniper]

I am not privy to the difference between the OEM and boxed versions. All I see is the difference in warranty. It could be possible that the OEM processors are generally offerings in a scratch and dent sale. I simply don't know. This is my first build since the old 486 days and I have not kept up. OEM processors are targeted at premade systems that the owner most probably will not tinker with, hence, a 57C temperature on one core at stock speed at 100% load is well with in Intel's tolerances. This processor's Core 0 will run all day long at 57C I am sure.

My backround: US Army retired instrument (test equipment) repairman and calibration technician (E7/SFC). Currently I am a Biomedical Eng Tech. I am also an Amateur Radio Operator. I've had some dealings with electronics for a few years. I made an attempt to get some really good parts (Mobo, PSU, RAM, CPU etc). Maybe I should have asked a little more on the differences between packaged units.

Thanks everyone. I hope more folks will chime in... maybe with a resolution.

Blessings to all,
ms

(yes I am)

 

[imposter]

I've read here that some of the E8400's have a problem with the temp sensor diodes on the chip getting "stuck", but that once your temps get past that point they start reading. But since I don't have an E8400 I can't confirm this. I do know that my 45nm Yorkfield has no problems such as some of you guys are getting with the E8400 procs. But it might have other problems with Core Temp setting the Tj Max too high. There is about a 10 C difference between Core Temp and Real Temp, which corresponds to the Tj Max assumptions of the 2 programs.

Spoken with Tc, i there will most likely not be a solution for those batch of cpu's.

 

[Wrathre]

Well, I adjusted my heatsink's seating and I'm idling around 31C. Much better now.

 

[fritzman]

Mine idles around 32c both cores and is around 45-52c full load. I'm using a TRUE120 and although that will lead to lower temps all around, I doubt it evens out the cores. I think if you just reapply the tim and try to reseat the hs it may help. Try a different orientation of the hs if you can.

Same... I set an OEM up yesterday in P5W-DH board, sitting under a TR cooler that has a 92mm fan (unsure what particular model is is, but looks a lot like my TR 120 cooler in sig, just smaller) and it idles at 31-32 and did 16hrs Prime at a max temp of 50 degrees (using CoreTemp).

I'd say you have a seating issue.

edit... Well look at that... you sorted it while I was typing the (right) answer lol.

 

[blazed]

Many e8400s will not report acurate idle temperatures. First of all use RealTemp, do the stuck sensor test, good chance you have 1 or 2 for idle range. As the chips heat up they become unstuck and more acurate, which is why the load temps work, just use realtemp or -10C from coretemp.

 

[abang_apy]

Many e8400s will not report acurate idle temperatures. First of all use RealTemp, do the stuck sensor test, good chance you have 1 or 2 for idle range. As the chips heat up they become unstuck and more acurate, which is why the load temps work, just use realtemp or -10C from coretemp.

Maybe you should update your bios...

Regards,
Abang_apy

 

[TankGuys]

According to Intel, this is not a problem with the CPUs. It's an issue with immature BIOS/temp software reading improperly. We've taken returns on chips with this problem, yet when I pop them into our test machine, they work fine.

 

[Joeflyer]

According to Intel, this is not a problem with the CPUs. It's an issue with immature BIOS/temp software reading improperly. We've taken returns on chips with this problem, yet when I pop them into our test machine, they work fine.

I just installed the E8400 on my ASUS P5N32-E SLI mobo after updating to latest BIOS 1403. I also found that BIOS shows 43C at idle but the case CPU temp monitor shows 24C at idle. I have the Vigor Monsoon Lite CPU cooler and never had core temp above 32C at full load with the E6750.
With the E8400 the CPU doesn't even feel warm to the touch when idle, so I think there is something not correct with the temp indicators. Just to be on the safe side, I reseated the heatsink twice and had applied a VERY thin coating of Arctic Silver 5. This should not be the issue at all.

 

[Coreyhm1]

I have been going by the Tc temp (CPU) instead, and it seems to be accurate. I know it may, will read a few degrees lower than Tj, but its accurate. The 74c max intel specifies, is for Tc, not Tj. I don't see any reason why not to go by this temperature.

 

[mormonsniper]

According to Intel, this is not a problem with the CPUs. It's an issue with immature BIOS/temp software reading improperly. We've taken returns on chips with this problem, yet when I pop them into our test machine, they work fine.

Thanks Tankguys for chiming in. I will do and learn all I can. I will do a 180 on the heat sink first and see if that helps. Making a 90 degree change is not what I really want to do as installing the back plate was a pain for my old eyes and hands. The CPU seems to be working fine, except that when I got home this afternoon, I got a white screen with the microsoft circle and a logo popping up, so I reset it and all is back to normal....

 

[mormonsniper]

Ok, did the 180 degree change on the Ultima heat sink. Temps are Core 0
56C and Core 1 is 41C, so no change. Next, I will attempt to update the bios.
TMPIN0 Temp is 34C and TMPIN1 is 30C (a little better as I remember).