DFI Admits 5Volt Vdimm Problem

[BB_One]

???

Ummm, puzzling. Could it be that some versions of the board are affected others are not ?

Had been running my VX for 3 months @ 3.35 with no issue until recently one stick now failing...cannot say if it is board related or just bad luck with a memory stick ~ frankly hard press to point the finger in either direction honestly.

Running my TCCD for now, and when time permits I shall RMA my VX and try again with new set I guess.

Given the shipment of DFI boards ( roughly 200 000 units per months) x the number of time I have seen sad comments about a dead or malfunctioning board ( in this forum and others). Assume a manufacturing defect rate of <3% which I believe is what DFI shots for on the LanParty series, as oppose to <0.98% on their commercial serie.

Quick maths for 6 months, all causes combined (users causes included), anticipated amount of defective boards ~ 36 000 units. Given the vast majority of forums I survey and review, my gut feeling is that DFI is meeting it's Quality control target, and nothing out of ordinary as surfaced to date, or we would have had an up roar similar to what was heard with the infamous ASUS board last year.

Dunno, honestly dunno...wait and see what comes out in weeks to come I guess.

Edit: Small note to topboy. I certainly respect the efforts you bring forward to share your findings, however you should probably be carefull with such Title as DFI admitting to something. Many will argue that your phone conversation, is literally hear say from one person to another one, irrelevant of who was at the other end of the phone. I am willing to bet that if you where to ask from the same person to send you the very same information by email, I suspect that it will come out differently, and might not support your claim.

You have I am sure heard of the famous, I heard from a friend that he heard his cousin heard from his aunt that she said she overheard her dad say that his sister had heard that an uncle had said that I was not the friend of my friend!

Wake up I told him, you are not only my friend, you'r my brother have the same mother and WE have a darn yappy family !

[Faiakes]

So what does this news mean exactly?

If I run my RAM at 3.6 Volt will I be ok?

[shimmishim]

i'm using the modified setting where you run the voltage off the 3.3 volt rail but have voltage choices up to 4 volts by tricking the board...

is this safe? or will this still cause problems?

[AndyT]

I appreciate your putting this out there topboy.
I saw the grief ya got from some fanboys elsewhere.
Travis just closed that thread saying there was already another..
There isn't.

Man DFI employees really irk me

Just a few level headed thoughts:

There is a thread about the "cold boot" issue that I think Travis is reffering to. People are posting on both topics, as they might be related in that thread.

This was probably bad timing for all concerned as most of the DFI decision makers are attending Computex in Tawain this week, through next. Travis is the Lone Ranger over there and is not the board designer, so getting any answers from DFI will take time I am sure.

OCZ and DFI have always worked closely together and we will help DFI as much as possible. I don't think anyone knows ecactly why the 5v jumper is killing memory, or why it is killing mostly CH5 UTT memory. TCCD/TCC5 is realatively unaffected, even when run off the 5v jumper setup on the DFI NF4, as well as most BH5 based memory (although a few kits had passed on). I still say the jury is out until someone can pinpoint exactly the cause (voltage spikes, too much current, dirty power, memory slot issues, ??). Personally , I never jump to conclusions.

I would suggest getting off the 5v jumper for now. The OCZ DDR Booster works great, as I decided to test it in the DFI NF4 as a possible sloution for those that do not want to crank up the 3.3v line and have a PSU with a -5v line (white wire in PSU mobo connector). Using the booster I was able to duplicate the 3200 Gold (BH die) timings and speeds I was getting using the 5v jumper.

Lets all take a deep breath and hope an answer will be available soon.

[AndyT]

i'm using the modified setting where you run the voltage off the 3.3 volt rail but have voltage choices up to 4 volts by tricking the board...

is this safe? or will this still cause problems?

The 3.3v Mod or "Trick" has pretty well been proven safe. I would not worry.

[AndyT]

So what does this news mean exactly?

If I run my RAM at 3.6 Volt will I be ok?

Read my post, 2 posts above.

[AndyT]

I hope Frank doesn't get the can for telling ya that ( I don't even know him).
I've seen folks get shown the door for less.

Me too.

[Snafu]

One thing we should remember is that this is a public forum and anyone either as an individual or as a corporation can come on line and provide an alternative position. ON some forums where this info is posted this has happened. On another forum the thread was shut down with mention that it was already posted but no link provided ( ).

As long as the info is correct and is not subject of an NDA there should be no concern of anyone posting info gathered. In fact I think it is important to share this info with as many folks as possible so that they are at least warned of the potential issue so they can make informed decision on what they do.

Kudos to topboy for posting this and for anyone else who shared info from their experience.

It would be interesting to see what ram folks are running with the board, their settings, how long they have been running and whether all was fine or if they had a problem.

[grdh20]

Is it safe to use the 3.3v jumper trick to get more voltage?

[Mikki]

As Andy stated a couple posts above,

The 3.3v Mod or "Trick" has pretty well been proven safe.

As with all things, nothing is guaranteed. We all push our hardware as far as we're comfortable with. The best thing is always to research and then make your decision based on what you find.....

[topboy]

I have been using full 4v several times, fon't have problem at all.
Below DDR593 2-2-2-3 are all in 4v

Kevin i take it Oskar has sent you over to do some damage control.

OK i challenge you

Put your DFI board up on the 5 volt jumper, run your ram at 3.4 volts with some nice overclocking to the board and run Prime95 continuously for 2 weeks 24/7 then come back and tell me the results.

I bet you will come up with problems with the ram after this.

Topboy

[eva2000]

The OCZ DDR Booster works great, as I decided to test it in the DFI NF4 as a possible sloution for those that do not want to crank up the 3.3v line and have a PSU with a -5v line (white wire in PSU mobo connector). Using the booster I was able to duplicate the 3200 Gold (BH die) timings and speeds I was getting using the 5v jumper.

Lets all take a deep breath and hope an answer will be available soon.

is that the modded booster with components switched around or the stock booster ?

[harleybro]

Hey thanks for the headup topboy. I actually lost one of my mobos last weekend after a flash. Was having all sorts of problems. Killed my 4000+ chip also . Just saw a thread on XS asking to test a bios from WU that had been out for a week or more. The one that I believe killed my mobo. But due to the paint and insulation stuck to it guess RMA is out of the question.

Was wondering when OPB starting posting but just saw it's the first post. Nice to have the bandwidth guru around.

[glorfy]

One point I don't want to miss here, is what other manufacturer gives you the option to run your ram off the 5v rail?

Overclocking is the art of running components outside of there recommended tolerances. Much kudos is due to DFI for providing a board with the options that it does. If there are troubles along the way, well I'm sure they will solve them.

I for one thank, them for giving me the option of breaking stuff if I want to They don't need to hide problems as far as I'm concerned - just work to fix them, and acknowlege them.

[topboy]

Hey thanks for the headup topboy. I actually lost one of my mobos last weekend after a flash. Was having all sorts of problems. Killed my 4000+ chip also . Just saw a thread on XS asking to test a bios from WU that had been out for a week or more. The one that I believe killed my mobo. But due to the paint and insulation stuck to it guess RMA is out of the question.

Was wondering when OPB starting posting but just saw it's the first post. Nice to have the bandwidth guru around.

Hey harley was this the 510-2 bios from Oskar