Cold Boot? Memory Died? [email protected] = who you need to contact

[tjj]

uhm, why has this been removed from the sticky..

[red930]

If a thread is very important you can always save it in your favorites.

[dsumanik]

Hey guys, after having all these threads merged i decided to go back and read through from the start to see if i could put some pieces of the information together in a different way, and I noticed a few things I thought were interesting.

 

1) Regarding Rgone's post about the bleedinedge 5v faulty jumper announcment:

 

The cold boot issue, or no-post problem (I call it both so we dont confuse english terms with what the engineers in Taiwan use, as per rgone's post)

 

to quote him:

 

Well I am going to tell you how such shett gets heard. Notice I did not say, 'said', so pay attention.

 

Oh I paid attention rgone, and this is what i think.

 

 

If Frank Wong reported this:

 

Mr Wong basically informed me that DFI R&D in Taiwan know about this issue and have been testing this jumper setting with many different major ram modules. He told me that the 5 volt setting is causing and i quote "Major Issues"....

 

....Mr Wong also said "At this time they do not have a fix for this problem. Our R&D has also found that the Vdimm jumper when set to 5 volts can cause some of the motherboard chipsets to blow"

 

to topboy (paul North of bleedinedge)

 

Exactly what were the terms we have confused there, or the miscommunication?When he said "cause chipsets to blow", did he mean ram modules? Is that what the engineers call them?

 

Regardless, anything blowing on a motherboard is not healthy, is it?

 

What were the "major issues" he was referrring to, if they were not related to this cold boot, ....ahem....I mean no post problem? I think all of us here deserve a slightly better explanation than that, just for the frustration we are all experiencing alone.

 

2) There are uncomfirmed rumours of a new sli-dr layout revision., drastically different with only different amount of pci slots and the memory in a DIFFERENT LOCATION on the board.

 

http://www.dfi-street.com/forum/showthread...hlight=revision

 

Angry claims he and Rgone have heard nothing..however that does not mean it does not exist. Is this yet another miscommunication between cultures? A typo? Could you guys dig a little bit deeper about this? If it is untrue, no big deal...but if the board is being changed so substantially...why? Expecially the memory location?

 

I am not trying to get you guys pissed off here, but i think we deserve some answers and need to know the credibility of these reports.

 

I honestly dont have any proof to say for sure that the board is killing memory, or likewise that the UTT modules cant hack the voltage. I just know this problem, and many of us here, had it occur using these components and it it definately has to be one or the other or both. I didnt blab to bleedinedge about a 5v jumper issue, Mr. Wong did, by his own free will.

 

Bottom line is, did Mr. Wong say this or not? If so, why? What was he referring to?

Is there a new revision coming out that is substantially different? Is it related to this cold boot problem, or others that have been discovered?

 

Would it be possible for Mr. Wong to speak to us directly in this thread about his comments?

 

OK, Enough about the gossip flying around this situation, this is what i have to say about the real problem on hand:

 

 

2)Although there are reports here and there of users experiencing this problem NOT using utt modules, or the cold boot occuring off of just using a ddr booster, or while just on the 3v jumper......

 

the majority of us have OCZ VX, or other UTT based modules running >3.2vdimm on the high volt jumper.

 

Also, I personally have yet to hear of anyone else running these modules on other NF4 boards exhibiting these problems, if someone knows otherwise please let me know.

 

 

Our Common facts:

 

A) We all have damaged or malfunctioning ram as replacing the ram with new stuff cures the cold boot instantly. If the board was geberally faulty, or had been "blown" from to much juice etc as mr wong stated...then new ram wouldnt fix the problem.

 

However..if it is replaced with UTT and high voltage is applied, the problem returns.

 

B) We all have "one good stick"...that was likely damaged while in slot 2.

 

C) We experience cold boot symptoms from any complete power off situation.

 

D) We NEVER experience the problem during a warm boot.

 

E) Once running, performance is STABLE and VERY FAST.

 

F) Voltage hits our ram when the power is off immediately aftyer a cold boot. I first noticed this when using my original corsair pro with the led's, they were showing activity before startup..Ozsnoal also confirmed this with his voltmeter.

 

 

What this tells us:

 

A) The board works fine on other modules, while on the 3v jumper. Many people have reported being problem free using UTT on the 3v jumper and doing the sense mod. The DFI is stable and literally ERROR FREE in this situation. I know this personally first hand. I can overclock my generic ram, run dividers and voltages till the cows come home while on the 3.3 rail and nothing bad happens at all.

 

B) Something is happening to one of the sticks that isnt happening to the other. We all have "one good stick" My theory of why our problems seem to slowly degrade and eventually dont work period is that once the cold boot issue occurs the first time...we monkey around with the memory to get it going...and the only way to get it going is by putting the undamaged stick INTO SLOT 2. Then the cycle repeats...and guess what, you eventually wind up with 2 damaged sticks.

 

C & D) There is a difference between how the motherboard treats the memory, and/or boot sequence when cold an warm booting:

 

Thanks to Ozsnoal we discovered there is definately a difference in how voltage is applied on cold startup...however this isnt the whole story. He then got a ddr booster to correct this problem..and MADE SURE that the mem had "proper" voltage on boot.

 

But the cold boot persisted in this situation too.

 

What does that tell us?

 

That the initial voltage may have something to do with the problem, but not all of it. Applying high voltage right from the get go doesnt fix the problem...only booting off of one stick does, then adding the other.

 

We NEED to isolate the differences in how and why the mobo detects memory in this situation, but not initially with both sticks in.

 

E) As stated above we all experience no problems once up and running, performance is astounding and everything we originally hoped for. This tells us there is at least one stable, fully functioning state.

 

F) Why the heck is voltage hitting the ram prior to startup..is this when the mobo detects spd timings?

 

 

If we put all this info together this is what i suspect:

 

Although the very real possibility exists that the VX memory cant handle 3.5 vdimm for any extended period of time...there may also be a problem with when and how voltage is getting fed through slot 2.

 

There has to be a difference or we would all have "2 good sticks" or "2 dead sticks"

 

We definately know there is a difference between how the mobo responds during cold and war boot...the key to this whiole thing and a possible solution may be as simple as having the mobo behave the same way in both situations...we need to identify the differences OTHER than how voltage is apllied to narrow this down. The juice hitting the ram prior to startup could be related to this.

 

Also since the problem is not occuring for the overwheming majority of people on the 3.3 jumper, then there also must be a difference in how the mobo treats the ram while using the 5v rail. Obviously votlage is way higher, but perhaps something else too? Is the startup sequence identical for the board while on either jumper?

 

Also out of curiosities sake, and as well as it being the only option not yet explored:

 

How is voltage applied to the memory during or right before POWER OFF perhaps this is when a potentially lethal voltage spike is occuring? After all, when we completely power off is when the problem is happening/and first noticed....could someone check this?

 

 

All I can say is this: i REALLY hope when i recieve the new modules from OCZ that these new components are stable and working because then at least we have a definate working combination we could replace all RMA's future purchases on.

 

what an ugly mess.

 

This is becoming a Sherlock Holmes mystery

[Angry_Games]

1. i dont have a clue what Frank Wong said to anyone. He talked to Topboy, so only he and Topboy know. I doubt either would lie, but what he said to Topboy isn't really a concern because if you DID read back through the thread, you saw where Rgone explained the communication problem about cold boot on the NF2 and how no one else but he and I understood exactly what we were talking about when we said 'cold boot'.

 

Frank and the rest of the company (who are all chinese and native chinese speakers by the way) thought cold boot was something else.

 

Whatever Frank told someone is what Frank told him, and I do not have a clue exactly what it was. I have no messages, emails, phone calls, anything like that from any of the engineers or anyone else within DFI that states that this is a 'known problem' and will 'blow chipsets'. If Frank has some knowledge that we don't that that would truly be news to us since he is the guy who translates all the mandarin chinese stuff into english for us.

 

Angry claims he and Rgone have heard nothing..however that does not mean it does not exist. Is this yet another miscommunication between cultures? A typo? Could you guys dig a little bit deeper about this? If it is untrue, no big deal...but if the board is being changed so substantially...why? Expecially the memory location?

 

we (Rgone and I) dont deal in speculation, rumor, and leaked info. Our job is support. We are not in board design, engineering, sales, marketing, or anything. Purely support. We only know what we know from first-hand experience, from what we see daily in tech email and inf the forums, and what our colleagues at DFI report to us. We don't base anything from a website's leaked info, reviews, previews, trade-show pictures, marketing PR releases, any of that. It is not important to us what is coming down the pipe as we only are concerned with what is out in customer hands right now. Period.

 

I honestly dont have any proof to say for sure that the board is killing memory, or likewise that the UTT modules cant hack the voltage. I just know this problem, and many of us here, had it occur using these components and it it definately has to be one or the other or both. I didnt blab to bleedinedge about a 5v jumper issue, Mr. Wong did, by his own free will.

 

I dont have any evidence of it either. Again, Frank is a guy in the office that does support. He is not an engineer. He is not part of the TW lab for DFI that tests these things, nor is he an employee of OCZ that tests these things. Frank might have misspoken or didn't understand exactly what was being asked or maybe he knows something that Rgone and I don't. I don't know. I'll state again that I have no knowledge of any evidence that the board is killing memory, and i have never heard anything said internally or from one of our partners that alludes to there being a 'known problem'.

 

I am not the CEO, nor am I anything but the support guy. Same with Robert. We cannot speak for Frank nor anyone else in the company. We can only speak for ourselves.

 

 

 

 

 

the rest is as you say...a lot of speculation and guessing as to what the problem really is. As I have stated before, I do not know what the problem is, and I am not the board designer nor the engineer and have no idea what to do about. It isn't my job to fix it. My job is to alert those that can fix it that there might be a problem and it needs to be looked into. I have done my job faithfully and have gotten not only Oskar involved but the engineers at OCZ. If there is no answer from them then you can only assume that they have not found real evidence of anything yet and are still testing (and they were last time i talked to both OCZ and Oskar).

 

Rgone and I have no way of fixing this situation. If we did, then there wouldnt be a situation because it would be fixed. I have done all that I personally can to help get this situation resolved, but whatever is required at this stage is beyond my capabilities as I am not an engineer, electronics expert, memory engineer, nothing like that. I'm a guy that has a job to do (as is Rgone) and we are here to do our job. Doing our job is what we do. Oskar and OCZ are doing their job as far as I know.

 

I don't have any estimation on time. If and when a problem is finally narrowed down and there is an answer, it will be provided to us to provide to you. If and when that happens, it will happen. I dont have any better answer than that. So please don't demand or release your frustration on me or Rgone about this.

 

We are trying to get a resolution for you and everyone else experiencing this. But we are not involved in the actual testing and fixing because it is beyond us what needs to be done.

 

if you are not satisfied with this answer, I don't honestly know what to tell you. It is the only answer I have for you.

[dsumanik]

now that is just not warranted at all...please refrain from such. your frustration is warranted, but your language nor attacking attitude is not going to be tolerated.

[Angry_Games]

i'd be for reading the above a little and take a bit of time to cool off before posting again...

[dsumanik]

Jeez,

 

angry I wasnt trying to attack you or rgone or anything?

 

I did read rgone's post please dont impl;y I didnt...why do you think i was asking the question of how Mr. Wongs statements became confused because of curtural differences?

 

I am not saying you or rgone, or Oskar or any of the OCZ guys arent doing a good job...at all....many, many companies would have already pulled out by now.

 

Not to mention we all appreciate having features like 4v vdimm in the first place, or memtest via the bios, or cmos reloaded.

 

This board kicks butt.

 

We just need to find the cause of this and move on, im doing everything I can personally to help and I am sure you guys are too.

 

I just was trying to get some answers, i guess there are none.

 

crappy. but lets keep going on with the show i guess, hopefully the new components work properly and dont die.

[OzSnoal]

Nice summary dsumanik.

 

New bios here, I haven't tried it - since I have gotten rid of my VX. Some people post say that it fixes their cold boot problems. Guess the bios must be here on DFI forums too. Add the link if you find it.

 

http://xtremesystems.org/forums/showthread.php?t=66060

[cabby]

Regardless of all the different opinions going on, the issue is not resolved and I would appreciate this being a sticky and not buried at the bottom of the forums, this is not a 1 or 2 user problem. I check everyday hoping that we hear some word and now I am digging thru the forums

[red930]

Once again...

 

If a thread is very important you can always save it in your favorites.

 

What's important to you might not be important to another forum member. Simply save a Favorite to the thread and you can reach it right from your browser. This way no matter where the tread is located in the fourm you will always be able to find it.

 

Regardless of all the different opinions going on, the issue is not resolved and I would appreciate this being a sticky and not buried at the bottom of the forums, this is not a 1 or 2 user problem. I check everyday hoping that we hear some word and now I am digging thru the forums

[OzSnoal]

You've heard the saying 'Out of sight out of mind'.

 

Still, this issue is not resolved and until it is or someone says we give up, it should remain stickied. It's all fine for the bookmarking and all, but what about the new guy that comes along? Then there is going to be more threads about cold boot, or will admins merge them like they did at the start of all this?

 

I suppose whoever unstickied it had their reasons.

[stewgy]

I was and still am a new guy and i agree with oznol, this should remain stickeid to benifit new NF4 LP owners when they first visit this site either in frustration panic or just to say hi, I know how much it benifited me when I first got here and when my board started exhibiting "cold boot sysmptoms" ohh and by the way I have never used the 5v rail. although had it not been for this sticky my situation might be much worse than it is. I am fully understanding of DFI's position, I know thier a good company and will do what is right, but none the less the people/customers have a right to know if their$$$$$ big dollar $$$$$machine that they just built may or may not have an issue. just my 2cents.