Why is my E8400 so hot??

Hey,

Alright, Ill start off by saying, that Ive done 3 re-mounts, with different AS5 / MX-2 Applications, and all 3 have yielded me the same results. Almost the same anyways.

Im running a DTek Fuzion with the 4.5mm Nozzle (Earlier I was running the Quad Nozzle, as I was using my Quad).. Temp's have improved by 4C, just because of that.

Ambients : 24-25C

My Load temp's are ~63/63C ... Prime takes it upto 65/65.

Last night, when the Ambients were 21C, Temp's were 61/61C Prime Loaded.

http://img441.imageshack.us/img441/4375/helpqo3.png

Only thing I can think of is that either coretemp is wrong, or the VID coretemp shows is correct.

I tried testing 333x9 @ the lowest vCore I could (1.200 BIOS Set, 1.110v Loaded Prime), It was stable for 1 hour, then I shut it, since I wanted to OC :P). Temp's during that run, were like 55/55C Prime Load.

Im running rock solid stable, Prime completed 9 hours this morning @ 4.3Ghz @ 1.39vCore Load.

Oh, Contact is perfect as well...

What could be the problem?

My E8400 is at 42C-44C at idle under water (stock). Under load it's 52C. Not sure what to make of it at the moment.

Mine is also as hot as yours. ~40C idle and ~67C load @4300MHz 1,42v under water.

They are hot little chips. I have tried 3 of them, under Tuniq all idle around 40 and high 50 while running prime with only 1.325v in bios.

Those temps seem pretty typical of many of the water cooled E8400's, with many load temps in the 60's and 70's...which seem surprisingly high to me.

Cooling, or even voltage increases seems to have little effect on Wolfdale temp readings. Some guys have removed their cpu fans completely and temps go up only slightly. I get the same cpu/core temps with the cpu fan on 800rpm, 1600 or 2250rpm...or with the voltage at 1.2v or 1.6v, makes no difference according to CoreTemp & CPUID Hardware Monitor. Intel's TAT does not work with Wolfdales.

My E6850 was 31C idle and 44C-48C load at 3.6GHz. But that was using the P5W DH Deluxe which only has 4-pin 12V connnector (not sure if that's why).

Hmm.. Well the E6850 isnt the same as a E84/85...

Im just surprised that the temp's are so high. I know Intel has said that the max. safe temp is 75C. But still, Id rather have it below 50C :P.

can someone remove their E8400's IHS? :)

Quote:

---

Originally Posted by Bail_w

can someone remove their E8400's IHS? :)

---

Thanks for volunteering.

awaiting result.

I'm experiencing temps around 65-70C on the Intel QX heatsink with my E8400 around 4400MHz, 1.5v. Haven't tested lower volts yet, just got the chip about 7 hours ago and I've been running benchies for a review. On the next heatsink in line(Scythe Ninja) I'm sure temps will drop 10 degrees to only be countered by raising the Mhz to 4500 or so. Got 4275MHz stable on the aluminum E8400 cooler, 4365 on the Q6600 heatsink, and now 4400 on the QX heatsink. Can't wait to put this chip under the cascade this weekend.

Perhaps the heatspreader isn't enought straigth.
Lap it and tell the results.

I wonder how on earth did all the reviews come out with the idea of E8400's giving lower temps. It's either the chip doing terribly hot or the software is not reporting correctly. I had my E8400 on IP35 PRO and Foxconn Mars but both shows high temps.

My E8500 reads the same temp at stock and with the cpu volt @ 1.5v with core temp so dont worry.

i have my e8400 running at 4230mhz at 1.368v vcore.
under water, apogee gt, idle at 42c and 68c under load...as5 used.
guess that is pretty high.
i think i didnt install them right.

Just tested my old E6600 and my new E8500.

6600 reports idle of around 34 to 37 degs.

E8500 around 43 to 45 (bios however says 29)

My E8400 Temps SUCK

i installed a TRUE to replace my Zalman 9500 and the temps are exactly the same! 38/42 idle 60-65 load at 1.34v

i dont understand it

Temp sensors can't read temps right. Try everest instead of coretemp.

my E8400 idles around 30° and has not gone over 43° even for a second under full load in my warm house. maybe you guys should back off the volts a bit.

Quote:

---

Originally Posted by s0lid

Temp sensors can't read temps right. Try everest instead of coretemp.

---

Everest, speedfan, coretemp and Dual core center all report about the same

It's because o the volts mate. Anything above 1.33v and they heat up like crazy. I have around 66C load @ 1.48v on air only but my VID is higher then yours. At 1.6v i hit it up with some orthos and temps jump straight to 75-83C region.... so you can see how hot these things get when applying serious voltage to them, compared to the Conroes or even the Kentsfields.

Ooh, and by the way, the lower the VID the CPU has (vid shown in coretemp), the higher the temperatures you'll get when giving it some voltage and upping the mhz.

Quote:

---

Originally Posted by Pandamonia

My E8400 Temps SUCK

i installed a TRUE to replace my Zalman 9500 and the temps are exactly the same! 38/42 idle 60-65 load at 1.34v

i dont understand it

---

I put 2 fan on the TRUE
at 1.43v Idle 35C/37C _ Prime95 60C/58C

I agree that droping the volts to a near minimum but stable level should help but perhaps the CPU is just going to run hot like the old P4 3.0+ Ghz did...

has anyone tried remounting yet?

Quote:

---

Originally Posted by AriciU

Ooh, and by the way, the lower the VID the CPU has (vid shown in coretemp), the higher the temperatures you'll get when giving it some voltage and upping the mhz.

---

Like I've said elsewhere, think about it. 1.6 vcore minus 1.3 VID = 0.3v and 0.3/1.3 = 23% overvolt. But 1.6 vcore minus 1.0 VID = 0.6v and 0.6/1.0 = 60% overvolt. Imagine overvolting the whole mobo by 60%, you think it'll get hot? Or catch fire lol.

But they are just asking for more and more volts. I think I buy another new fan rather than lower the volts :)

^^^

i'd suggest moving to a nordic country for lower temps but you're already there :)

move the PC outside :clap:

to those suprised at the temps, look at the low VID of the chips ....it's a gimme when you're pumping 1.4+ through.

I thought 1.45 was the max to give these CPU's, as they are 45nm and 1.45Vcore corresponds to 1.55-1.6 for the older generation... You guys are really pushing these cpu's too far for 24/7 use... just giving them too much volts...

I will have an 780i and E8400 next week will see what gives there... under water and air, but the CPU has to be used fully under air, I'll see where I can find the sweetspot between mhz, Vcore and temp...

Quote:

---

Originally Posted by Leeghoofd

I'll see where I can find the sweetspot between mhz, Vcore and temp...

---

and isn't that what it's all about! :up:

All new 45nm CPU´s are hot after 1.4V barrier.

I have a 19 degree difference between my chips at idle (45 and 26). Do you think these temp monitoring programs are wrong or one core is messed up?

Well it seems my QC9650 is borked too, idles at aroudn 45-44-39-40 and goes up to 50-50-48-47 at stock speeds, if I up to 3.6ghz 1.25 volts it gives the same temps lol... I can hardly trust temp readouts and I think it's a general thing with all new 45nm... I was shocked to see those temps as I would have liked it to be 10°C cooler than my Q6600...bump

Quote:

---

Originally Posted by Leeghoofd

Well it seems my QC9650 is borked too, idles at aroudn 45-44-39-40 and goes up to 50-50-48-47 at stock speeds, if I up to 3.6ghz 1.25 volts it gives the same temps lol... I can hardly trust temp readouts and I think it's a general thing with all new 45nm... I was shocked to see those temps as I would have liked it to be 10°C cooler than my Q6600...bump

---

Hey m8, finally u got your QX9650 :up:, I got a E8500 here that I am going to test 2morrow.

load temps seem to make some sense on mine. Idle temps make little sense. I have remounted once (though was sure it was good first time), and tightened and loosened mounting screws while running with no effect.

stock clock 1.12v, idle cpu 24, core 47. So core is 23C hotter than tcase at idle at 1.12v...mmmmm ok.

stock clock 1.12v, load cpu 32, core 50. So core is 18C hotter than tcase at load at 1.12...maybe.

Now does anyone believe 23C delta at idle and then decreases to 18C delta at load? mmmm ok :rolleyes: .

OC 4.1 and 1.38v (and 1.42v as well) idle cpu 29, core 49.
so increase volts 23% and increased idle 2C mmmm ok.

load made sense though was 66C

Let me join the boat guys, I've got the same issues as everyone else regrading E8400 and temps! I've run it on my Maximus Extreme and my DFI P35 with same results. At idle with all settings stock (3.0Ghz @ 1.08v) my temps are 422C-46C under decent watercooling. Under load (4 instances of Prime @ Small FFT's) the temps only goes up to 48C-53C. Everest Ultimate and CoreTemp are responsible for those temp readings above. If I run the Asus PC Probe II or the Smartguardian on the DFI the temps are much more resonable at 24C-26C (Idle) and 33C-37C (Load) at same settings as above in both boards respectivly.

I should also mention, although its probably not relevent to anything, my CoreTemp VID level has changed a couple times (and not becaues of board changes). When I first installed it into my DFI the VID was 1.100v, but a couple hours later (don't know exaclty how long it had been, wasn't paying attention to it until I noticed the change) it read and still reads 0.9750v? It's a bit discouraging I guess because I'm affraid to really push it or do stability testing overnight because I'm not sure what is really going on here (are temps supposed to be much less overall with new Fab or what)?

Mine idles at 49c and hits mid 50s under load at 3ghz 1.1v on a p5e-vm.
Reseating made no difference.

By the way guys RGone just recently posted some very relevent information regarding this E8400/E8500/QX9650 temp problems over in another thread. HERE is the link to his post which has links included. Fortunetly it looks like the issue is with Intel and not motherboards, BIOS versions, or CoreTemp as these things are doing what they are supposed to be doing!

50C load is about right on decent air cooling at stock 3GHz 1.12V if the core probes are working.

has anyone took the time to remount to see if temps decrease? This time try placing thermal paste in a thin line across the center of the CPU like this:

http://i11.photobucket.com/albums/a1...S5_Install.jpg


All these threads should be consolidated into one main sticky thread (I hope).

XS thread
XS thread

Yes Eastcoasthandle, I've re-seated my chip multiple times and I always apply TIM in the manner you suggested! Over the years I've learned that is not only the easiest way to apply TIM but also gives the best results (compared to smearing it all out yourself, like I used to do)! Unfortunetly the multiple re-seats did nothing to effect temps in anyway!

^^ Ive tried 4 Remounts now, with different applications, including the above.

OK well it's time to send email.

Sounds like a good idea but I wonder if you actually get an answer from anyone?

You won't know until you try

Quote:

---

Originally Posted by Roger_D25

Sounds like a good idea but I wonder if you actually get an answer from anyone?

---

Prob the same answer for everyone who will e-mail them regard about the E8400 heat issues.

I got a hot one too...
http://img219.imageshack.us/img219/1561/84065386au4.jpg

ambient temp 24-25C

8400 4ghz@1.33 idle 42C load 65C or so.. on water, rig in sig..
remounted the lil bastard, so the contact is good. my e6600 was running colder hah

read this post

I just re-set my wb. Same as before..
http://img159.imageshack.us/img159/9732/42602214il2.jpg

okay some people might think i am b.sing, but so far intel is replacing my chip the e8400 - i got this to replace my e6850 (bad move) - anyway

i tried the stock cooler, and the intel one that came with the cpu, much smaller and lighter than the e6850 and my thermaltake v1.

i dont even have enough time to load core temp, but in the bios

it reached 105C four times and shut off. does that damage the CPU?

anyway when i pulled the heatsink off, it seems that the coolers dont make any contact with the cpu.. the stock heatsink had the thermal spread perfectly still on the fan. and the article silver on the cpu with the v1, never touched the v1 base.

So the two heatsinks i tried can't make contac twith the cpu,properly.
i crapped when i saw the temp reach so high in the bios, i thought for sure was a bios bug, but nope.

Quote:

---

Originally Posted by Ad1tya

Hmm.. Well the E6850 isnt the same as a E84/85...

Im just surprised that the temp's are so high. I know Intel has said that the max. safe temp is 75C. But still, Id rather have it below 50C :P.

---

Imho, there are some misleading speculations about temperatures of 45nm. CPUs. Lets put the things in order.
1. There are some key words : Tcase, Tcore and TCC (Thermal Control Circuit).
2. According to Intel, Tcase is:
For the processor, the case temperature is defined as the temperature measured at the geometric center of the package on the surface of the IHS.
It should not exceed 72.4 grad. for E8400.
3. Tcore is temperature measured by DTS which is placed in hottest spot in the chip. So naturally Tcore > Tcase.
4.TCC:
The Thermal Control Circuit portion of the Thermal Monitor must be enabled for the processor to operate within specifications. The Thermal Monitor’s TCC, when active, will attempt to lower the processor temperature by reducing the processor power consumption.
The primary function of the PROCHOT# signal is to provide an external indication that the processor has reached the TCC activation temperature. While PROCHOT# is asserted, the TCC will be activated. Assertion of the PROCHOT# signal is independentof any register settings within the processor. It is asserted any time the processor die temperature reaches the trip point.
PROCHOT# can be configured via BIOS as an output or bi-directional signal. As an output, PROCHOT# will go active when the processor temperature of either core reaches the TCC activation temperature. As an input, assertion of PROCHOT# will activate the TCC for both cores. The TCC will remain active until the system deasserts PROCHOT# The temperature at which the PROCHOT# signal goes active is individually calibrated during manufacturing. Once configured, the processor temperature at which the PROCHOT# signal is asserted is not re-configurable.
5. DTS. According to Intel:
The processor uses the Digital Thermal Sensor (DTS) as the on-die sensor to use for fan speed control (FSC). The DTS is monitoring the same sensor that activates the TCC (see Section 4.2.2). Readings from the DTS are relative to the activation of the TCC. The DTS value where TCC activation occurs is 0 (zero).
Conclusion from 4. and 5. - there is no direct way to get actual core temperature. We may calculate it by assuming some TCC activation temp (Core Temp and other progs does it) but it is not 100% correct.
On the other hand the actual core temp is meaningless - we don't know the specs of Intel's silicone. So the only important thing is to keep DTS readings less then zero.

This is all nice theoretical stuff but we still don't have a clue then how hot the cores are, the developpers need to get proper input from intel to get their software reading out the cores correctly or at least be able to give us an idea if we are still in the safe zone... but thing that worries me the most is that some get let us say real values while others are really messed up: too much difference between the cores, idling at sub ambient temps,.... so there might be more here going on then we can imagine...

Time to call Mulder and Scully !! and maybe Mc Guyver too

anyone tried a temp probe or gun and compared to their older E6 series?

I've got the hardware to both use a temp-probe or gun to measure temps but don't know how to properly take the measurements? It sounds easy but how do you go about using say a "K-Type" probe or non-contact probe to measure the CPU temps? I thought about maybe drilling a small hole perpendicular to the bottom of my CPU waterblock and mounting my temp-probe into the hole using some TIM to fill any voids? Although that might work somewhat it certainly won't be accurate because of the material between probe and CPU, plus I don't know how deep I'd be able to drill into my CPU block (making some type of fixture to hold block so I could drill using mill or drillpress would be difficult at best? If anyone has any decent ideas please share them!

1)underclock and undervolt cpu and be quick or hold your block against and remove and measure, but if undervolt enough, ok with nothing (when later increasing back to normal, increase volts first, save, then increase mhz...or enjoy reboot hell.)
2)take block off and put piece of masking tape over back of cpu (temp is inaccurate with IR gun off shiny back of cpu without tape...learned that from unclewebb post)
3)start computer
4)I positioned screen so that cpu temps and IR gun both visible in pic....temp 2 is cpu temp.

this is my E6850. The cpu (tcase) temps were very accurate in software (temp 2), comparing reading to laserpro temp gun (calibrated accurately with very accurate temp probe).
Note the core temps are off 15C, as this test was done before it was known that tjmax was 100c, so should read 75 and 74C.

I have done the same with my E8400, cpu temp reports same as IR gun at 50C and 60C. Need two people for pics, so no pics of E8400. Who knows about core temps.

btw...when i had looked this last year...tcase sensors that listed as cpu temp are supposedly fairly accurate (always exceptions/bad sensors), within few C, especially between 50 and 80C, where they are calibrated for...listed in some chart in intel specs...too long ago to remember...and to lazy search again for.

http://www.xtremesystems.org/forums/...d.php?t=179044

real temp program WORKS FLAWLESSLY stock e8400 ubder water now reads right 34-30