DFI Admits 5Volt Vdimm Problem

[harleybro]

Yeah it was. The one I saw on XS. Since I have having problems getting my ram stable w/ a divider gave it a go. I can't for sure blame the bios on the chip. But I cut the ihs off and it worked great. Hit 3gig stable on stock vcore. Went into bios and was having problems getting a divider stable. Next was flashing the bios. Afterwords couldn't boot the chip set bk to stock bios kept locking. Threw the chip on another mobo and smelled smoke. Threw a 3500+ on the same mobo and it works fine. Maybe the bios flash killed it maybe not. Got a FX-55 yesterday and grabbed a brand new mobo (just in case) No flashing now until I know it's ok. Time for some FX goodness on ocing this weekend.

[Old_Fart]

This is my first post and I hope I have something useful to contribute.

If you take a look at my sig, you'll see what I'm running. I've had "cold boot" issues since I built this system and have been corresponding with Robert (AKA RGone) at DFI support. Here is a direct quote from an email I just received from him this morning ---

"I was trying some VX4000 day before yesterday and it will not boot windows solid above the 240x10 mark at all. Did not have much time to fool with it as I don't care much for memory that only performs once you put a ton of voltage on it and get again 'more' heat in an already hot system. It was such a pain to get the VX4000 tweaked and since day before yesterday I have re-configured the board 7 times for other testing and may get back to the VX later."

He went on to say ---

"... there is some research being done to see if there truly is a problem or if it is underpowered power supplies or what. It seems only Hi Voltage UTT type memory and using the 4.0Volt jumpering makes a problem. But it is still unknown for sure."

I'm wondering now if the cold boot issue is related to the jumper used to increase DDR voltage to the board. I guess only time will tell and hopefully there will be a solution forthcoming.

I'm fairly certain that my power supply is NOT the issue (see sig) so it must be either the ram or the board or the combination of the two (along with possibly the memory controller in the Winchester core). I suspect it's the latter but I also understand that OCZ and DFI worked very closely on this combo.

I've also read some rather glowing reviews on the DFI Ultra-D and OCZ VX ram combination, which is why I bought it in the first place. I'm already considering a CPU upgrade to the new San Diego core (A64 4000+) and have read that the new memory controller in these chips is a big improvement over all the other A64 cores.

I guess it also doesn't help much that the documentation on the BIOS settings in the DFI NF4 boards is so pathetically bad (non-existent?). One would think that based soley on it's level of complexity, that DFI might shed some light on this subject OFFICIALLY. But no, it appears we're all stuck with having to weed through endless forum posts trying to find our exact combination of parts and BIOS settings that MIGHT work (if we're lucky).

Having been in the [computer] business for 30 years now, I've seen vast improvements on the hardware side of the equation but the software and support end of things only continues to worsen. Back in the OLD days it was very common to pick up a telepohone (does anyone out there remember what those are?) and in a matter of minutes be talking to an actual ENGINEER who was involved in the design and implimentation of the product you were calling about. What do you suppose the odds of that happening today are?

At any rate, I'm not going to turn this into a rant. Lord knows we've read enough of that tripe around the net. I will say though that I hope DFI and OCZ can figure out what's going on and solve these problems quickly so we can get back to actually USING our computers for something other than test-beds and benchmark machines (just to find out if we're actually getting the performance we PAID for!).

Later,
Old Fart

[timpanogos]

Well, not much, but some input from a dfi-streeter ...

http://www.xtremesystems.org/forums/...4&page=7&pp=25

[topboy]

!!!!!NOTE!!!!!

To all

Anyone who is having problems with the 5volt jumper could you please also post over at this thread please

http://www.dfi-street.com/forum/showthread.php?t=12342

This may make the "Tech Support" wake up to our problem

Thanks

Topboy

[topboy]

AG at DFI-street just posted an official response to the problem

Looks like we are in trouble here as they are saying there is no problem

Quote.......

To conclude, there has been no evidence that the DFI NF4 motherboard has been responsible for any memory death. You can choose to not believe this if that is what you want to do. We have done extensive testing and found no evidence that the board kills memory modules though, and that is the truth, no matter how much some of you might want to flame and shout that you 'know for sure' that the board does kill memory.

Post link......

http://www.dfi-street.com/forum/show...2&page=6&pp=15

Looks like we need to get onto the memory vendors about this

Topboy

[pastorjay]

Well, I think there will be some angy memory vendors now! THis thing may take a while to play out.

[dsumanik]

My board and pc4000VX is now being sent to oscar wu himself to get to the bottom of this, ill keep everyone posted.

[Mikki]

It seems that Oskar Wu and DFI are working hard to find out what's goin on with this and find a solution, and I know that there are a lot of others involved in the research and testing of this issue. Whatever the outcome, it's great to see so many people around the Net working together on this. Kudos to everyone for that, once again we see that the enthusiast community is a great one to be involved in.

I'd also like to mention that BE's involvement in bringing this issue to the forefront is completely in the interest of us...the enthusiast community. BE has had a good relationship with DFI in the past and we hope to continue that. We've been impressed with their boards, as have the rest of the enthusiast community, ever since the original LanParty boards....I personally own 3 boards, including the subject of this thread. We've always tried to work with manufacturers to solve problems and make their products better, we're proud of that.

The important thing is the community is now aware, and now we just need to be patient and hope that a solution, whatever that is, can be found.

Here's to the enthusiasts around the Net....

[Snafu]

[muzz]

I am extremely impressed that Oskar himself is not only looking into this, but letting folks know his findings as well.
This is one of the reasons he is a great engineer.
Heres to hoping for a resolution to this issue.

m

[gijoe]

idd i think ive had the same problem
http://www.bleedinedge.com/forum/showthread.php?t=10364
i am sending this baby back.
sad joe

[WxChaser]

Mikki and all -

I've yet to have an issue with the higher 5V Vdimm voltage jumpers being applied on the SLI-DR or the Ultra-D mobos using 3200VX, OCZ PC3200 Gold, or Redline PC4000 Redline - YMMV!!!

I believe that bigtoe and others have examined the voltage spikes in more detail/depth than I've had time to spend in my curiosity - and believe the method of
pumping up the 3.3v rail instead using adjustable pots or the 3v sense mod on a PSU is perhaps a better way of getting around the unstable voltages that could be possibly damaging equipment.

Using a Powerstream 600W PSU I have not noted big Vdimm voltage spikes on the SLI-DR @ 3.5-3.6v yet with my DMM (with 5v Vdimm jumper enabled on the mobo).

[topboy]

Where this maybe a solution for some, it might not be feasable to others.

The next text is not directed at any one person, just my thoughts on the issues.

I think we need to look at the issues here.

1. DFI NF4 boards are killing RAM on the 4 volt jumper FACT. Ask AndyT or anyone from OCZ.

2. DFI have admitted that there is a problem, some believe this, some dont. FACT is there were 2 other people that heard my questions and reaction while on the phone to DFI.

3. DFI sell these boards stating that it is capable and supported to run a 4 volt jumper. OCZ sell RAM that is capable and supported to run RAM at 3.4 volts (VX RAM).
SO if i have a board that i bought that is advertised and supported to feed my ram UPTO 4 volts and i buy RAM that is advertised and supported to do 3.4 volts, how come people including me are killing there ram at 3.2 volts. This is where i argue back and say i want it sorting, i paid the money to the companies, it doesnt work, fix it or give me back my money.

And i dont feel this is an unreasnoble thing to ask, and i think alot of users feel the same.

When i get my new board back and exchange my 3200 ram im going to test this for another month. If i see no fix in this, then im onto the companies for a refund in full. Failing that it will be off to the lawyer and the office of fair trading. Done it before, will do it again.

Sorry if im ranting but im really hacked off with compaines supplying goods that do not do what they say on the label.

Topboy

[Alexo]

Does this problem happen all kinds of overvolted memory or are certain types more susceptible?

Does BH5 RAM die as well?

[topboy]

what volts you running alexo