Jump to content

Question about Expert and dying CPUs


Recommended Posts

the moderators haven't attempted to assist him yet... or anyone with a dead cpu from the expert board for that matter

 

 

It's hard to assist know-it-all's who belittle you and everyone else before they even get a word in to assist you. I'm sure YOU'LL be a lot different. :rolleyes:

 

BTW: This is really starting to sound like the High vDIMM low vCore warning Big Toe (AKA OCZ Tony) Already warned everyone about. The only reason this would happen more frequently on an Expert board in comparison to something from another manufacturer, is because the DFI board allows a higher vDIMM to be selected in the bios.

 

I believe Tony's rule of thumb was to have the vCore atleast @ 1.425 for any vDIMM above 3.0v. Remember, this more than likely wouldn't kill a chip right away, but would be a slow painful death to the memory controller. So even if you bring your computer back down to stock clocks, the Memory controller has already taken the harsh abuse. And actually lowering the vCore now may have lead to instability/death.

Share this post


Link to post
Share on other sites

  • Replies 106
  • Created
  • Last Reply

Top Posters In This Topic

dixevo, not to be disrespectful in any way, but this is not a thread to discuss something that has nothing to do with the problem at hand. Try posting something of this nature in the nForce4 or AMD Overclocking section of these forums. :)

 

it does have something to do with it. READ the post again. What i'm saying is that this might not be a problem with DFI per say. It might be a broader component issue.

Share this post


Link to post
Share on other sites

@computerpro3

 

At no time did AG confuse FSB for LDT. The only person that made a claim regarding "underclocking" and "LDT" was you .

 

Did you come to DFI-Street with a bad attitued and PWNage on your mind. Yes by your own admission.

Share this post


Link to post
Share on other sites

It's hard to help someone that comes to the forum to inflict PWNage.

 

"Grab some popcorn and watch me pwn his butt, I really don't care if I get banned. Lets see how long it takes for them to hide my issue."

http://www.xtremesystems.org/forums/showpo...80&postcount=15

 

Are you sure you still want to defend him?

the moderators haven't attempted to assist him yet... or anyone with a dead cpu from the expert board for that matter

Share this post


Link to post
Share on other sites

You have a point ExRoadie, his reason for coming here does not help his case.

 

After reading this post (http://www.dfi-street.com/forum/showpost.p...07&postcount=33) by jnick and seeing a fair amount of these sudden cpu deaths with the expert board I (and everyone else should) wonder why DFI continues to flat out deny that there may be an issue. And on top of that has not done any tests to either prove or disprove the issue so it can be laid to rest.

Share this post


Link to post
Share on other sites

Although the memory controller is on the CPU, does the voltage increase on the ram chips via vdimm also reside on the cpu side? Or am i missing something?

 

Is increasing the VDIMM on the RAM directly proportional to the Voltage on the CPU memory controller?

 

Or how does it work?

Share this post


Link to post
Share on other sites

The VTT is what gets to the memory controller on the CPU. VTT is derived by taking VDIMM and dividing it in half. Max JEDEC VDIMM is 2.85v so VTT is around 1.425v normally.

 

Take into account now that AMD's Cool'N'Quiet lowers VCore to 1.1v I believe? (Never used it). So we can say that the theoretical maximum difference between VCore and VTT is around 0.3v.

 

If the original poster was running 3.5v VDIMM, his VTT would 1.75. Factor in the 0.3v max difference and you get a safe VCore of around 1.45v. He was running 1.53v so this whole low VCore/High VDIMM should not be in question here.

Share this post


Link to post
Share on other sites

A

 

computerpro3 was pointing out that although he reached an overclock of greater than 30%, greater than 3 GHZ, it shouldn't kill a CPU, because the CPU wasn't receiving enough volts to fry it. This isn't EXACTLY true, but it's accepted as a general rule. If a CPU overclock isn't 100% stable, timing in the opening and closing of the logic gates in the transistors is off, and thus the chip is unstable, and will slowly be eaten away.

 

That "out of sync" is rubbish.

 

It s true however that the lifespan of an overclocked CPU is reduced. First by voltage, then by high frequency (even at stock volts) then by heat (as long as you stay under AMD's limit).

 

So overclocking without voltage increase still limits the lifespan of the CPU, througgh frequcncy and heat.

Share this post


Link to post
Share on other sites

computerpro3, welcome to my ignore list. I hope you find a comfty place and sorry you are alone, not many people make it there these days.

 

 

I also notice that he said he was running 3.5vdimm. I thought that many of us "experts", including the people at Xtremesystems knew that running a high vdimm and a low vcore killed Athlon 64's? Oh wait...maybe some people aren't as expert as they like to present themselves. I'm going to quote what Tony (from OCZ, the guy who made the -bt(a) BIOS's and knows what he's f'ing talking about) said:

 

 

 

That sounds like it could be his problem, no?

 

This is is very unlikely to be the problem that causes CPUs to die on Experts for the same users that did not have their CPUs die on the Ultra-D/SLI-DR series.

 

Since the memory controller is integrated into AMD64 CPUs and pretty much directly connected to the RAM there is really not much that the board can do to accelerate the above problem. It is unlikely the Expert makes the above situation worse or better than any other AMD64 board.

 

The new board could theoretically increase this problem by accidentally overvolting the RAM, but as you can see in the tables about theoretical limits (http://cracauer-forum.cons.org/forum/viewtopic.php?p=83#83) you would need a whole lot of volts in Vdimm to kill the CPU. If the board would do that then only Winbond chips would survive to tell the tale. Any other RAM would die earlier from Vdimm than it would kill the CPU.

 

Since we have only seen much less RAM deaths on experts than CPU deaths none of this is likely.

Share this post


Link to post
Share on other sites

Alritely then.... now for the opinion of a guy who knows little about about ocing computer (venice 3200s are progressively getting worse clocks these days :() but a fair bit about english since hes still @ skool (yes im makin an idiot out of myself here wit the grammar - but its midnight)

 

@ computerpro3 or woteva ur called. i think it would be fair to call you a fool rite about now. the number of times u have contradicted or incriminated urself is getting beyond funny and im kinda getting sick of it. to make a long story short - ull help ur case more if you stop posting!

 

@ anybody on the 'pros' side. okay he mightve had a point @ the start. but now is he here to find a solution or just try to bring down the credibility of DFI because hes that kinda loser? i think that question answers itself - and just for computerpro3 its called a r-h-e-t-o-r-i-c-a-l question!

 

@ the sane peoples side (inc a_g exroadie and rgone...)

i am so sorry u guys are sittin @ ur computers readin this. if he was within arms reach, somebody wouldve at least slapped that 'expert' in the face by now. Dont let him piss u off too much - the rest of us still love u - lol! Just leave him to destroy his own argument and get on with life

Share this post


Link to post
Share on other sites

Guest
This topic is now closed to further replies.
 Share


×
×
  • Create New...