I returned my 8400 today!!!!

[Thanatos421]

Another thing that seems to ring true in all cases on these posts. The Tcase sensor DOES seem to be reporting correctly. It's the DTS on the cores that seems to be new/buggy.

Download Everest and check the "Sensors" tab. Look at the CPU temperature, and then compare it to the core temperatures. It should be no greater than 10 +/- 4 delta. If it is, there is your proof that the Tjunction sensors are junk.

 

[sprucebr1]

One other thing, if you don't like the E8400. Try an E8200 or even a X2 5000+ BE.
I doubt this has anything to do with a $200+ CPU being completely defective.

 

[TechnologyCoordinator]

X2 5000+ BE.

That's the only AMD chip that can get decent overclocks, it's not too shabby in the low-mid dual core market.

PS - The Star Trek movie has been delayed by the writers' strike, it won't be out this year.

 

[jimmysmitty]

This is an example of trying to do something without first researching how. This CPU (E8400) is running at 3.0Ghz stock. WHY would you go so far as to try and get a 1000Mhz overclock out of this? 3.0Ghz is an amazing frequency to begin with. Point is, the E8400 is fine at 3.0Ghz, when installed properly. There's no reason to push this to 4Ghz or more. The stepping won't make a difference in this matter. Intel is not to blame for someone's uneducated mistake.

Only reason to have it run at 4GHz is because that is even more amazing on air with little extra heat. That is the fastest a chip has ever been OC'ed on air. Before 4GHz+ would nee better cooling.

Its kinda like why OC a Q6600 to 3GHz. Just b/c you can with ease.

 

[caamsa]

TC just fast enough? Who would ever say anything as silly as that

S939 baby!

It's a little pokey, but I simply can't justify the cost of upgrading when it's, dare I say this, fast enough.

Whoops....................got ya! :lol:

 

[zenmaster]

Its a defective cpu period, your claims that we should just ignore the problem shows your bias towards intel. The fact of the matter is we paid good money for a fully functioning cpu, not one with bad sensors and bad HSF's etc. Amd did the same thing with the phenom, should we just ignore the problem because the errata will only occur in a rare virtualization??? This is about helping the consumer get what he paid for, not any fan boyism. Im just simply suprised that Intel would do something like this after all the fall out from the phenom fiasco.

Maybe if you did not have an "AMD" logo in your sig and your reports jived with what other people are reporting on the Web, folks might believe your rant.

Example - 0 Folks on New Egg reporting problems. Sure they are all not that bright, but you would think one person would have had an issue :>

Or perhaps somebody else here would have had that issue.
Has anyone ever seen that issue - Sure.
Maybe it was Mobo Related maybe not.

However, if the issue was as rampant as you claim others would be posting it heavily.

 

[xnamerxx]

Well we have a client with a pentium d 830 that does alot of video rendering. There has been almost nothing I can do to keep that system from overheating. I was able to keep the processor at a nice cool 74c with a aftermarket heatsink and the northbridge was a chilly 63c. Now because this was a preschott processor I cant complain but I dont think your little 8400 is running hot at all.

 

[funnyman06]

NO, NO, NO....the supposed heat issues of the E8xxx series boil down to bad temperature sensors!!! PERIOD! Noobs don't know how to use a digital thermometer that you can pick up at Radio Shack for 15 bucks and trust a sensor saying they're hitting 80c. It's a bad sensor, you didn't need to return the CPU, you just needed to find accurate temperature measurements.

If my temp sensor that was built into my CPU showed 80C all the time id return it also. When i overclock my CPU i want to know how hot the core is getting. Not the heat sink, the core. No 15 dollar temp prob is going to tell you that. It might get close but its not the core temp. Plus i run a water cooled rig, so the core temp is the only temp i can pull from. Regardless how fast it is stock, its always faster when overclocked.

We had absolutely no problems cooling an E8400
with the ASUS V-60 Vapo Bearing HSF:

We used a very modern thermal sensor: MY HAND!

The much smaller size of Intel's stock HSF
should also tell you something (HINT HINT).

Then again, if you throw your E8400 into a food blender,
and it gets scratched a lot from whirling around at 7,200 rpm,
I wouldn't complain to Intel about that.

But, whiners are in the habit of shifting blame whenever possible
-- probably public school emissions .. er .. graduates!

Also, i have a P4 3.0E Prescott core, even when i feel its heatsink it is not ever hot, but i know the CPU is running somewhere around 50-60C idle. So the hand is a horrible way to test things.

 

[Thanatos421]

OK, big deal, you take the Tcase and add no more than 20c depending on your cooler, BAM there's your core.... Have people lost all ability to think critically and extrapolate their own data??? Or do all you sheep depend on what your sensors tell you as 100% concrete.

 

[funnyman06]

OK, big deal, you take the Tcase and add no more than 20c depending on your cooler, BAM there's your core.... Have people lost all ability to think critically and extrapolate their own data??? Or do all you sheep depend on what your sensors tell you as 100% concrete.

Well some of us, when we are OCing, have temp ranges we want our CPU's to be in. You cant possibly say that you are always going to be below this temp. This is far to general of a statement. Maybe this person was under the temp, but for me that is not acceptable.

Plus if im paying 300 bucks for a CPU, i really want something as simple as a stupid sensor to work correctly. Its not like this is something new. They have had it correct before and they should have it working now.

 

[Thanatos421]

Wrong, thermal sensors have always been flakey at best. I know for every PC I've ever built and overclocked, I've had to calibrate my thermal monitoring before I would ever trust it 100%.

So this is NOT new. Now instead of adding 15c, you may have to subtract 50c....big deal.

 

[KyleSTL]

Bigger better faster more.

Here's the new rig I'm putting together.

Celeron 440 2.0Ghz (OC 3.6)
Asus Striker II Formula
Corsair DDR2 1111 Dominator 2x1GB
2 Palit 8500GT 1GB SLi (awww YEAH)
6 WD Raptor X in RAID 0 (OS)
3 F1 1TB in RAID 0
Thermaltake SwordM
PC P&C 1200W

It's gonna roxxx your soxxx.

LOL. JK. n00bs are fun to make fun of.
Seriously, it's Friday and it's been a long week. I think I'm becoming delerious.

 

[caamsa]

Maybe if you did not have an "AMD" logo in your sig and your reports jived with what other people are reporting on the Web, folks might believe your rant.

Example - 0 Folks on New Egg reporting problems. Sure they are all not that bright, but you would think one person would have had an issue :>

Or perhaps somebody else here would have had that issue.
Has anyone ever seen that issue - Sure.
Maybe it was Mobo Related maybe not.

However, if the issue was as rampant as you claim others would be posting it heavily.

I did an exact phrase search on google "E8400 heat problems" and got 59 hits. Of course some of the hits I am sure are repeats. But it is being talked about so there must be some issue.

Here are two of the links.

http://www.pureoverclock.com/story1791.html

http://www.legionhardware.com/Board/viewtopic.php?t=3378

 

[Grimmy]

Heh.. all he (eternal) had to do was return it for another. And keep doing that till he got one the worked properly. That might have been a hassle, but I would see people doing the same for the phenom chip.

But now this guy had changed his avatar to AMD, and just took it off, since he claims this is not a fanboyism thread, and has, or seems to have a brand new spanking 6400+ system. Yeee haaaa!!

He could have just got a Q6600, and got more accurate temps and complain how high they get.

I've said that Intel and AMD are the same. They both want your money.. and attention. See now AMD has hurt Intels feelings for getting all the bug attention.. now it's their turn. 😗

 

[funnyman06]

So this is NOT new. Now instead of adding 15c, you may have to subtract 50c....big deal.

And that i find unacceptable...

 

[pausert20]

I got to ask the question. Who and were has it been proven that the Digital Diode is reading wrong? Anyone even think to question the software is not interpreting the reading from the Diode Correctly? Remember that the thermal diode is not even supposed to be fused to be working on the Wolfdale processors. Intel made the decision that the Digital Diode will be the way to measure core temperatures using the PECI interface bus.

 

[LukeBird]

But Luke, aren't you supposed to come to Intel's defense with the same arguments of "overclocking isn't important" and the famous, "but is it fast enough?"??

jkjk, but seriously, jkjk

He he, not worried I'm going to cry in the corner at those comments... 😉
Well yes, I did make such a comment! 😀
Although as I've also said, for me OCing is something I don't have any interest in. Although I am getting interested in it and am considering 790FX, 3870X2's in CF and a Phenom BE. Although the last bit is the only bit I can't make my mind up on! :lol:
And other than 3DMark where I'm CPU limited, my CPU is more than fast enough thanks! 😛

 

[funnyman06]

I got to ask the question. Who and were has it been proven that the Digital Diode is reading wrong? Anyone even think to question the software is not interpreting the reading from the Diode Correctly? Remember that the thermal diode is not even supposed to be fused to be working on the Wolfdale processors. Intel made the decision that the Digital Diode will be the way to measure core temperatures using the PECI interface bus.

I don't own an E8400, so i couldn't tell you, but if i went into the bios and it showed 60C+ temps all the time, then something isn't right. I don't know if he is using core temp, or speedfan or what.
If it is reading correctly in the bios then i would agree the software isnt mature enough to understand the processor.

 

[Grimmy]

I don't own an E8400, so i couldn't tell you, but if i went into the bios and it showed 60C+ temps all the time, then something isn't right. I don't know if he is using core temp, or speedfan or what.
If it is reading correctly in the bios then i would agree the software isnt mature enough to understand the processor.

Soo.. your using the "What if" thing?

Can you link this from where ever.. or does this happen to do with someone you know?

 

[Thanatos421]

Pausert20 - the point is, whichever sensor is reporting the core temps is way outta whack. None of these people seems to think that hey...the Tcase is only 50c, and my HSF is only mildly warm to the touch, so hmmm...no way my core temps are in the upper 80s....duh.

 

[funnyman06]

Dude, read what he said. I can only extrapolate from his posts. He says it has heat problems, someone else says its a temp prob problem. I don't know, i don't own one, but giving them the benefit of the doubt, id say he has the most experience with the E8400, being the only person here who has had one. Also, the store where he returned it said to him they have gotten a lot back, so its not an isolated incident. That is all.

 

[Grimmy]

Heh.. the other thing about that is when people get the Tcase temp from once source (bios, or MB app or Speed Fan) and then see another app that shows Tjuction temps, then they really go crazy>>>>>> . o O (help me.. I'm melting.. I'mmm..mmmeeettttiinnggg)

😳. o O (I'll go sit in the corner for awhile)

 

[buzznut]

Well, if we continue to accept thermal sensors that don't work properly, then the manufacturers will continue to put out faulty products. One way of showing disapproval is by sending the product back. I don't see the problem with that.

 

[Thanatos421]

No, it's not an isolated incident. All the people returning them because they are "hot" are the same people who have no business building and overclocking in the first place. At the first sign of something funny happening (which always happens) they panic and RMA. System building has become so easy as of late that the majority of first timers and semi-newbs have NO troubleshooting ability. So they return a perfectly good part, and whine about it on a forum somewhere. I would hate to have seen some of these same people build an old AT machine. No color coding, plugs that could fit almost anywhere....I have a feeling we'd see a lot of posters magically dissapear after they crossed their ground wire for the mobo connector 😛

 

[trevorvdw]

So if you buy a new Lexus and the brakes work but the brake lights don't and its a flaw that Lexus knew about and didn't fix its no biggie because people can see you're slowing down anyhow or you could go buy some aftermarket lights and strap them to your trunk. Its outrageous to want to have something work as it should and you shouldn't bother taking the car back to Lexus and have them fix it, just live with it .. it works "well enough" ... lol