I returned my 8400 today!!!!

Page 6 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
LOL there are no 'heat issues' with Wolfdales. I've seen the horde claim this, do they realise Wolfdale is the coolest running desktop CPU available, with power consumption of around 30W under full load?

What it *does* have is an inaccurate temperate diode. You know what, it's not the first one to be inaccurate either. My Northwood P4 2.6C was reporting idle temps at or below room temperature, which we all know is impossible. From what I've read, A64s and X2 series also generally underreports temps, in fact a lot of older CPUs did this as far as I know.

The only difference this time with the E8x00 is that the error is on the other end of the scale, and it's reporting temps higher than reality. This leads to the horde crying 'OMG its overheating' when its just the temp being reported wrongly.

A faulty temp diode does not make the CPU run at 100C, the same way a faulty thermometer does not make you have a fever. 😉

If you are interested in seeing what the 'real' operating temps are like on the E8x00, check this link:
http://www.xbitlabs.com/articles/cpu/display/intel-wolfdale_11.html#sect0
 


:lol:. o O (I suppose it takes one to know one)

But for the record, you didn't provide any real proof you had one. All you had was your word. If you didn't know how to post pic's you could have asked. I bet allot of the guys here would have looked at what you were seeing to try to help.

If you going to dish out criticism, be prepare to receive criticism.

The temps you did say really weren't high at all in your other thread.

Anyhoo... have fun with your 6400+ system. If it really is a 6400+ system.
 
Being an AMD fan boy now running an E6750 @ 3.6GHZ and a Yorkfield @ 3.8 GHZ I can truely say that there is no reason to return something becuase it can not do that which it was never supposed to do well. This is not even true, becuase the E8400 overclocks well, but the thermal temps are wrong. Has the original poster said the CPU crashed, or had any from of defects? No. Just becuase an inaccurate thermal monitor was higher than you liked, is not a justified reason for returning it. This is not even the true case here, the CPU is yours to do what you want with, so you returned it, no big deal that was your choice. Making a very biased thread on a well respected forum about how terrible something the E8400 is, that is where you went wrong and now you are to blame for your mistake, so do not take offence when somone calls you on it. This thread is right in one sense the E8400 has a slightly exaggerated version of the defect in all CPUs, so maybe we should just not buy CPUs since they have problems and because it is just not fair.
 


Hell yeah its a justified reason, thats why it got returned.

Also I just saw the price go down on the q6600 go down to $199 so I might get that although I still think I should wait for the new phenoms and or the 45nm cuad core. My problem is I want a cpu that runs at 65w and the q6600 amd the 6400+ I currently have do not. They both suck up electricity and thats what Im trying to get away from and keep high speeds, and no bunk temperature readings of course.
 
Last night I did a lot of research on this because, unlike most of the people whining about a major defect, I actually own a e8400. What I found was this-

The only temp sensors on a wolfdale are the two dts tjunction sensors on the cores. These are the ones read by coretemp and some other programs as core 0 and core 1. These give readings as the difference below tjmax ( the point where themal throttling kicks in). Intel wont tell what the value for tjmax is, but testing has shown it to be around 105 c.


The cpu temp read by most utilities and bios' is coming from a sensor on the mobo, near or under the chip. If this reading is wrong, as mine is, blame the mobo manufacturer or software writer.


Many of the dts sensors are sticking at low temps. They will show a min temp for one or both cores above the actual. Both sensors on my chip work fine.

Intel doesnt consider this a defect. They have stated the dts's are not to be used for temp reference, their only pupose is to turn on thermal throttling. so if it reads correctly at high temps, it is fine. Perhaps this is why the mobo manufacturer utilities dont use it to read temps, because they werent intended for that. While many are incorrect at lower temps, I couldnt find a single report of one not working at higher temps. There may be some out there, but they are rare.

So while it would be nice if the tjunction sensors read correctly at lower temps, who cares what temp it idles at anyway, as long as it doesnt overheat under load. I would like my cpu to cook me breakfast, but Im not going to return it because it wont.


 


:heink: . o O (?)

And your other post:

he motherboard reads it correctly in bios and since so many of us just got it people havent adjusted the applications yet I think??? Anyone have any ideas because I would like to over clock with the correct temperature showing and core temp is incorrect.

Sorry to burst your bubble.. but you can't have both, low power and high speed.

If you going to OC, it's going to eat allot of electricity when the cores are under load. 😗
 
Well Grimmy a E8xxx or Q9xxx series will use less power than a Phenom especially at 45nm since they will have to set lower clock speeds. But you are right if you want high speeds/low power that ain't gonna happen anytime soon. Maybe with Nehalem.

But either way the OP is making too big of a deal. There are many other things showing how it doesn't truly effect anything.

Oh well. You didn't like it, too bad your loss.
 
I'm aware that the 45nm will use less power. But that is pretty much gone when you OC.

Even the Extreme calculator reflects that. If you go to it, to starts off with 38watts, so basically I choose the CPU at its stock, then do an OC and minus 38w.

Now it seems to be off a lil, since I chose my E4400 (2ghz) at stock with 100% load, it showed basically 55W instead of 65W.

It's basically the same with the E8400 (3ghz), 55W at stock, even though it's 65W. So that is quite an accomplishment at stock.

However those numbers do change in OC'ing. With my OC (3ghz, 1.325v), the calculator shows 94W. Same happens with the E8400 (4ghz 1.2v), but still impressive 72w. Once you really kick it up there with 1.4v 4.2ghz, its over 104 watts.

The Q6600 on that calulator shows (G0)82w (B3)91w stock, which would be still off since it suppose to be 95-105w.

So all I was pointing out that he wants 65w CPU, with best speed OC in mind.

Edit: I knew you agree.. but felt like making my keyboard babber some more. 😀
 
I swear, I've never seen Grimmy like this before! Eternal has managed to take one of the most laid back, silly, good-natured posters and unleashed the Grimm-fury!
 

:lol:

Err..

:fou: . o O <(Whooophhh!! *cough**cough**cough*)

😳. o O (sorry)

Edit:

:pt1cable: . o O(it's spring fever I tell ya... where is it!!)
 
Funny how all the fanboys are trying to justify Intel selling defective chips, yet they feed on the AMD Phenom hate. What comes around goes around.

Funny how you need to point out things that are important to you, not to everyone else.
 
Oh god here we go again..... MrsBytch, the CPU core of the 8400 is 100% fine! The entire chip may even be fine according to some who tout the BIOS updates and software updates fix this. This has 0 issues with performance.

Noboby in the know cares if a temperature sensor reports off, because they always have. Like I said before, and like Epsilon echoed, this time it's just reporting high instead of low and noobs everywhere are freaking out.

The really funny thing is, you didn't see these same tards freaking out before when they were actually burning up their chips because the temps were reporting LOWER than actual.

In my opinion, that is a much more dangerous ground to tread on. Then, you have a false sense of confidence when you think your chip is running cooler than it actually is. No one is paying that detail any attention tho. Most are just argueing for the sake of arguement.
 
Funny how all the fanboys are trying to justify Intel selling defective chips, yet they feed on the AMD Phenom hate. What comes around goes around.

Yeah, speak for yourself. You supported the devil, now AMD sucks.

You should have never betrayed AMD and got that E6300, its bad karma for all the AMD fanboys and fanbytchs. :lol:
 

Good point. But I never said any thing about voltage. I would say his voltage is about 1.2-1.33v. Also check out:
http://www.xbitlabs.com/articles/cpu/display/intel-wolfdale_12.html#sect0

maxover.png

Image by: xbitlabs
Imo, 70C is quite high but the OC is 4370Mhz. and its a E8500 so there are some differences.
 



hmmmmmm... Since the CPU report higher than real temperature, it should never kill it. It might trottle it down or even turn it off but not kill

LOL
 


I built old AT machines with no color coding and LIF sockets. That's how I learned to troubleshoot. Yet, when it comes to overclocking, I'm a noob because I only did it once for a short while with a Williamette. Yet, I still think that both Phenom bad core reports and Wolfie sensor and/or core temp issues should be investigated and not just argued over as if it's just "phanboyism" -- AMD fans gloating over Intel's similar issues.



I guess you don't know Fry's. Outpost.com does not have the weekly Fry's store specials. In the Friday morning paper is a Fry's insert with motherboard CPU/bundles. I wish I'd brought mine home from work yesterday. They had a Phenom with an ASUS board for $274, they also had a Conroe with an ECS board for $199. They had an X2 5600+ with an ECS board for $119. It varies week to week.

After I bought an MSI barebones with a 405 chipset board in February 2007, I got the same board bundled with an X2 3800+ for $179 the next week at Fry's. I've since switched both my X2's to AMD 690 chipset boards, but it was a good deal at the time.

Never rely on Fry's online store to list their in store specials. They aren't Newegg, they're a brick and mortar store. That said, I never liked their ECS boards, but have gotten CPU's with an MSI or ASUS board before and it turned out okay. His problems could be the ECS board's bios not being up to date to fix the sensor issue.

Now, where did I hear people kvetching about how lousy AMD was in requiring a bios fix for the B2 errata? I heard it here from Intel boosters. Well, if a bios fix is fine for an Intel board with sensor issues, then how can they kvetch about the bios fix for the B2?



Don't rely on that bad core report regarding AMD, it's a small number that must have gotten through QA, if it's genuinely due to a bad core and not overclocking issues. My guess is it's a bad core on some Phenom's that should have been binned as triple, but that's just a guess. Don't run with guesses just to diss the other company. That's fanboyism.

Both need to be investigated and not just argued over on message boards. I do agree it's probably a motherboard issue, especially if it's a reconditioned ECS board in a Fry's bundle. Still, it's an issue associated with an Intel chip that needs investigation and correction. Just as AMD was flamed for not getting all their AM2 board partners to get the bio's updated to support Phenom, so too does Intel have a responsibility to get the bios fix in place with all vendors. Fry's has some responsibility too.
 


Don't recycle. It needs to go in the local landfill. The reason is that when we recycle, the old PC's and monitors end up in China or India where no tech workers "recover" gold etc. from operations just outside their little bungalow's door. Most children of families who do this have lead poisoning.

We're told we're saving the planet, but we are hurting our brothers and sisters in India and China. Please read the following:

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1913570

Besides, archeologists 10,000 years hence need to find some of our junk after our civilization is totally destroyed by an asteroid impact in 2050! :na:



If it was an AMD processor, he'd be getting accolades. As is, he's being accused of being an AMD fanboy (probably true) by Intel fanboys.

The other group that's criticizing him aren't fanboys, but are enthusiast overclockers who warn that noobs can't overclock and assume he's a noob based on what he said on another thread.

Me, I think that both AMD and Intel have faulty processors that don't overclock well (AMD), or do not report proper temps when overclocking, but that run well at stock (Intel). Stock is virtually a banned word on these forums. If the DT's on the Wolfies are for throttling, then what are overclockers supposed to use to determine temps?

IMHO, for most people, overclocking is not worth the performance. It's a hobby that some enjoy. It's not something that everyone should do, even if they can figure out how.
 

And you had to post here that you returned a chip? Nobody gives a crap what you returned. Maybe I should post that my dad returned his power supply, because it was for a justified reason.
 
This post is like sitting in a fifth grade girls locker room. Full grown men bytching at each other like little girls. returning a fully functional (as advertised) piece of hardware for nothing is the most foolish thing I can think of. Guys like that are the reason good new hardware stock dwindles, and prices go up. Get a grip, people.
 


That reminds me of the boycott of soccer balls made by children in india. The media got wind of the fact that little children were exploited by having them sew those balls. After a successfull campaign at boycotting the balls made there, the indian industry adjusted. The kids that didn't become unemployed ended up mining stone and others resources with quite a death toll, while some of the young girls ended as prostitutes. It didn't take long until people realized that sewing balls, while not perfect, was one of the best options.

By saving those kids from elevated lead in their blood stream and poisoning the western worlds ground with perfectly recyclable material, let's just see what nice and healthy alternatives capitalism will provide them with.
 

Quite true. Intel never advertises anywhere on the box that you can read accurate temperatures, therefore there is no reason why you should expect to be able to. You can epect it to run at the rated speed and have x amount of cache, that's about it.
 


I edited your post and corrected all the typo's you made 😉
 



LMAO. Nice. Thats a good one.

yipsl, I never stated all of them were bad. Just that there are some out there and AMD needs to take action and either get them back or replace them with a fully functional chip.

The E8400 not being able to read the temps right has a workaround in the OS and I am sure Intel is working on a BIOS fix to it.

I think that either company needs to fix the problem but for me Intels diode sensor seeing higher temps as compared to the chance of getting a chip with a bad CPU is not as big a deal. I think that I would rather have my temps being seen higher than they are than have a bad core that can cause Vista(both XP and Vista have the hardest time BSoDing) to BSoD.
 


Well... Your right, I don't know fry's, but then what you say does makes sense. It's just the site even list out of stock items. Why wouldn't the E8400 not be listed? Even the phenom 9600 is listed, 5600+.

But I can see perhaps a really cheap MB combo'd up with the E8400 for a quick sale. Would have been nice if he could list is spec's to begin with, which I normally try to tell people, but sometimes just get sick of being a broken record player. :pt1cable: