Is my E8400 overheating?

[systemlord]

42- 45 is average reading on that chip your okay, mine idles at at 42-43 even though I think its really more like 28 -30 since people are overclocking these to 4.0ghz.

You have not a clue to what you are talking about, lol. Most of these people are either using a massive CPU cooler or water cooling. Stock cooling is for stock clocks, don't ever think your temps are cooler than your being told by temp programs! Not everyone is using the same cooler.

 

[systemlord]

Here's the thing though. Something, somewhere is reporting the temperatures incorrectly.

I have the E8400 also at stock speeds and with the stock cooler.

My Asus Suite and Speedfan both say idle temps are 28-30C and under load (running two Orthos for ~15 min) they show temps at about 45-47C.

CoreTemp shows idle temps at 45-47C and under load, temps were 61-63C.

Not sure which, if any, are accurate. I remember reading that these 8400's had something funny with the thermal sensor not working right. Not sure if it was just a theory or if it's been proven or not (haven't read anything new about it). So I'm guessing either Speedfan/MoBo software is reporting wrong temps or CoreTemp is.

PS. Bioshock is amazing. 😉

Intel processors have two different DTS (digital thermal sensors) one is measured inside each core and the other is measuring on die temperatures, on die temperatures are 15C degrees cooler than core temperatures.

CoreTemp shows idle temps at 45-47C

These temperatures are core temps.

Speedfan both say idle temps are 28-30C

These look like CPU temps or on die temps, see the 15 degrees celsius difference? Here is the temp guide you must read through.
http://www.tomshardware.com/forum/221745-29-core-quad-temperature-guide

 

[sources]

Okay, I'm thinking of RMA'ing it. Problem is I lost the plastic wrap to put the cpu/fan in. I have the actual box it came in, with the name in it and the barcode and everything, but the plastic thing it was sealed in is lost. Also, the pushpins are a little bent. That's obviously not whats wrong here though. Is it still possible to RMA through newegg/intel?

 

[Eternal]

^^ Your fine just get either the Zalman or the Coolmaster I posted. Both use screws that even the cooler perfectly over the cpu, that way you dont stress. But Rma away if you think that another chip will solve your plastic screw problem cuz it wont those plastic screws just plain suck. Its already been posted that the sensor on these chips might be off and they are probably running cooler, than the aps are displaying. I tried 2 coolers on mine just to make sure and the temepratures where practically the same with both, but when I used the stock cooler I noticed it was higher because the plastic screws where not staying in correctly, causing the cooler to rise above the cpu slightly so that the cooler was not even on top of the cpu.

 

[systemlord]

Just buy a better CPU cooler problem solved, those plastic screws just suck.

 

[randomizer]

Why do you people have so many problems with the stock heatsink mounting system? It's so simple and I've never had it pop out.

@OP: If it isn't throttling or shutting down, it isn't too hot.

 

[hughyhunter]

Jeeze... just get a new computer!

 

[Thanatos421]

OMG, how many times does it have to be said before you will listen. All the other posters are ignoring the fact that the E8xxx series procs have faulty temperature sensors, OR immature BIOS that is keeping your CPU from reporting properly. I promise you, you are not reaching those temperatures. Get a digital thermometer and aim it at the bottom of the heatsink. That's the only way to truely tell your temps at the moment with the Wolfies. I don't know if the same fault exists with the Yorkies, so I won't comment.

 

[sources]

Alright, thanks Thanatos. I'll do something like that.

Anyways, I called up intel and they said that I should be fine sending it back as long as the heatsink and cpu is back, and I shouldnt worry about the pins. But he said with my problem he doesn't see an issue with the chip and I wouldn't be able to get a new one because it seems to be fine, high temps on cores or not. He said that the general CPU temps is what should be monitored and that if I have an issue with a core temp then the general cpu temp would be a lot higher. My CPU is at 33 degrees C so he said that was completely normal, and that he cant see it being RMA'd due to the cores. But he said in the future if there is an issue then I can RMA it.

Is he right? Should I only pay attention to the general CPU temp?

 

[systemlord]

Why do you people guess whats going on? Why dont you download coretemp, TAT, and speedfan and compare the three. Likely you will find two that are similiar and one program thats way off. Then you will know what your temps really are instead of guessing.
Damn Intel for shipping bugged processors. 😱

Thats what you get sometimes for buying the first batch of any new product, theres almost always a problem with some graphics card makers at launch. Welcome to the cutting edge of pushed, rushed out new products to market. MrsBytch you seem suprised, this has been going on for long time now, don't be suprised anymore. O yea Dam AMD for releasing those buggy slow performing Phenoms. 😀

 

[sources]

Well, I REALLY want to OC this to 3.6GHZ, but I don't know if I should. I should get a digital thermometer, right? Would anyone like to recommend one?

 

[Zorg]

How many times, and from how many sources, do you have to hear it for it to sink in. The core DTS sensors on the 8400 are buggy. Try skipping the digital thermometer and put that money into a high end heatsink with a backplate. Even if you determine that the CPU isn't running hot, which you will, you can't overclock very far with that POS stock heatsink. Especially since you have installed and removed and installed and bent the push pin and it won't stay in anyway. Sheesh.

Take my entire post as though it was done in all capitals and in red.

Try listening to the people that respond to your questions.

 

[CompuTronix]

Thank you Zorg. For the 3rd time, sources, do yourself and all of us a favor; read the Temperature Guide: http://www.tomshardware.com/forum/221745-29-core-quad-temperature-guide

 

[cd14]

I am surprised nobody here mentioned this already--

The 8400's are known to have bad heat sensors. Heck, mine says it is 55-65C at idle as well... No biggy. It is running cooler then your apps will say.

+1

There's reports all over the net about "stuck" temps and overboard temp readings. Looks like programs are having issues reading the info from the CPU. This is all on different motherboards too.

I use coretemp and have gotten the temps to 104c (wouldn't go any more). It doesn't throttle or shutdown and my heat sink is just warm to the touch. At idle my readings are 59/63 at any speed (3 to 4ghz). At stock 3.0, the temps don't move at all when stessing or at idle.

I have a Q6600 which I pushed >80c and I can tell the difference just by measuring the temp of the air in and out of the case.

 

[Shadow703793]

@Mods:
Can some one please make a sticky on the E8xxx temp sense problem?

 

[CompuTronix]

Shadow703793, I'm presently working on including this problem in Section 15 (Troubleshooting) of the Core 2 Quad and Duo Temperature Guide: http://www.tomshardware.com/forum/221745-29-core-quad-temperature-guide

The faulty temperature sensor issue is not confined to the E8000 series processors, but is also common to the Q9000, and QX9000 series processors.

Comp

 

[Zorg]

That's great, Thanks Comp.

 

[sources]

How many times, and from how many sources, do you have to hear it for it to sink in. The core DTS sensors on the 8400 are buggy. Try skipping the digital thermometer and put that money into a high end heatsink with a backplate. Even if you determine that the CPU isn't running hot, which you will, you can't overclock very far with that POS stock heatsink. Especially since you have installed and removed and installed and bent the push pin and it won't stay in anyway. Sheesh.

Take my entire post as though it was done in all capitals and in red.

Try listening to the people that respond to your questions.

I said it before that I bought the Arctic Freezer 7 Cooler...

 

[hughyhunter]

This topic is dead... you can overclock... just dont go above 1.4v and 1800fsb to be safe!

 

[randomizer]

Shadow703793, I'm presently working on including this problem in Section 15 (Troubleshooting) of the Core 2 Quad and Duo Temperature Guide: http://www.tomshardware.com/forum/221745-29-core-quad-temperature-guide

The faulty temperature sensor issue is not confined to the E8000 series processors, but is also common to the Q9000, and QX9000 series processors.

Comp

An issue perhaps related to the transition to 45nm? I would have thought that intel could have found a bug like this by now since they've been working on 45nm for a long time.

 

[Zorg]

I said it before that I bought the Arctic Freezer 7 Cooler...

Good, no pun intended on the term sources. So I'm sure you understand by now that there are problems with the sensors, so read Comps temp guide Section 15 (Troubleshooting) when he gets it done. I suspect you will have to ignore the core temps until it gets worked out. This isn't the first time this sort of thing has happened, and it surely won't be the last.

It appears that the Arctic freezer 7 uses the same garbage push pin design that is on the stock heatsink. You need to get a heatsink that uses a backplate. It will require you to remove the mobo to install it, but it will be time very well spent.

 

[Zorg]

An issue perhaps related to the transition to 45nm? I would have thought that intel could have found a bug like this by now since they've been working on 45nm for a long time.

Who knows, maybe they were tired of everyone buying the cheap CPUs and OCing the he!! out of them. Although, if you look around it hasn't really seemed to put much of a damper on the OCing.

 

[randomizer]

JKflipflop said Intel are supposed to be more agressively binning their chips now. But I think the route of higher FSB, lower multi is a cheaper and faster shortcut to the same result. Look how poorly the Q9300 does. It also requires quite a high vcore to reach 3.5GHz as opposed to, say, a QX9650, which indicates differences in binning.

 

[Zorg]

With any (a lot of) luck AMD will put some pressure on them again.

 

[hughyhunter]

JKflipflop said Intel are supposed to be more agressively binning their chips now. But I think the route of higher FSB, lower multi is a cheaper and faster shortcut to the same result. Look how poorly the Q9300 does. It also requires quite a high vcore to reach 3.5GHz as opposed to, say, a QX9650, which indicates differences in binning.

I'm thinking the lower multi might not make much of a difference when we see FSB increase to 1800... that will allow a very high FSB overclock... if 1800 is standard then 2000 plus wont be so hard to achieve.