DFI NF3 Ultra out?

[[XC] DragonOrta]

It looks like Atacom.com has the DFI NF3 ultra in stock. Has anyone ordered one so that we can make sure they are shipping them? DFI NF3 Ultra (http://www.atacom.com/program/print_html_new.cgi?&USER_ID=www&cart_id=443420_152_163_101_10&Item_code=MBA9_DFIX_LP_24)

[Fission]

The store I am waiting for is NCIX, they are supposed to have them tonight/tomorrow, but still waiting to see. I will try and order one soon here, but as of 5pm PST, still no stock. Hopefully by the weekend, but if they are actually shipping, that is exciting enough for me.

[terror_adagio]

They start shipping tomorrow.

[LuckMan212]

sorry to be such an idiot but can I ask, why is DFI releasing an NF3 board after they already have NF4 boards out for awhile... is there some advantage to the NF3?
sorry if its a dumb question

[THunDA]

sorry to be such an idiot but can I ask, why is DFI releasing an NF3 board after they already have NF4 boards out for awhile... is there some advantage to the NF3?
sorry if its a dumb question


Its going to be AGP.. So this is great for people that arent ready to upgrade to PCI-E..

:toast:

[Dissolved]

think ill wait till newegg gets it..

might be worth the buy, if nto i may wait for the shappire mobo.

[Absolute_0]

I'm interested. My Neo2 has been acting incredibly buggy lately and i'm trying to get an RMA from Monarch. Been one week so far... :slapass:

[conrad.maranan]

Would you guys be interested in a review of the DFI board? If yes, is there anything in particular that you would like to see tested?

[Absolute_0]

Would you guys be interested in a review of the DFI board? If yes, is there anything in particular that you would like to see tested?

I'm just curious about the overclocking ability and the Vdimm potential. I'd love to see a review comparing it to the Neo2 involving max CPU speed, max HTT and memory overclocking comparisons.

[conrad.maranan]

Test Bed #1
DFI LANParty UT nF3 Ultra-D
AMD Athlon 64 3500+ (Newcastle)
Sapphire Radeon X850XT Platinum Edition
OCZ EL DDR PC-3200 Gold VX (2x512MB)
OCZ PowerStream 600W Power Supply Unit
Western Digital Raptor 74GB SATA HDD

Test Bed #2
MSI K8N Neo2 Platinum
AMD Athlon 64 3500+ (Newcastle)
Sapphire Radeon X850XT Platinum Edition
OCZ EL DDR PC-3200 Gold VX (2x512MB)
OCZ PowerStream 600W Power Supply Unit
Western Digital Raptor 74GB SATA HDD

How does that sound? ;)

[Absolute_0]

Be even better with a venice but i'd say it looks excellent! So you have one of these new boards? (or will soon? :))

EDIT: I'm curious, the memory is OCZ VX which requires volts... the Neo2 has a max Vdimm of 2.85 so how will you be doing memory comparisons? OCZ DDR booster or Vdimm mod on the Neo2?

[Dissolved]

Would you guys be interested in a review of the DFI board? If yes, is there anything in particular that you would like to see tested?


id rather you let me test it myself... you know you want to drive up beach and visit my lovely city ;)

[conrad.maranan]

Be even better with a venice but i'd say it looks excellent! So you have one of these new boards? (or will soon? :))

EDIT: I'm curious, the memory is OCZ VX which requires volts... the Neo2 has a max Vdimm of 2.85 so how will you be doing memory comparisons? OCZ DDR booster or Vdimm mod on the Neo2?
My Newcastle has a fairly decent memory controller. You may refer to this link (http://webpages.charter.net/conrad-digital/3500VX/10x263MHzROCKSOLID.png) for an image of what my Neo2/3500+/VX combo was able to do. The board will be shipped out tomorrow and I will review it as soon as I receive it. My opinions regarding the board will be published within a few days after initial testing. ;)

I will utilize an OCZ DDR Booster dialed in at 3.6V when testing the Neo2 board. The goal for the DFI will be to outclock the MSI unit's proven clocks of 10x263MHz 2-2-2-10-1T-1:1.

[Zebo]

Conard great! My question is what formation the ram is in for dual channel. Is it split or side by side like MSI and EPOX... I am having severe ram heat issues because they are essentially touching between the sticks.. I like how the NF4 board spreads ram out for better cooling and am wondering if NF3 does too. thanks.

[xenolith]

sorry to be such an idiot but can I ask, why is DFI releasing an NF3 board after they already have NF4 boards out for awhile... is there some advantage to the NF3?
sorry if its a dumb question

It's basically an NF4 with an AGP slot instead of two PCI-E/SLI slots. FYI, the pic @ atacom.com is of the NF4 Ultra-D. :lol2:

[Absolute_0]

Sounds good Conrad, i'm anxiously awaiting the review :)

[conrad.maranan]

Conard great! My question is what formation the ram is in for dual channel. Is it split or side by side like MSI and EPOX... I am having severe ram heat issues because they are essentially touching between the sticks.. I like how the NF4 board spreads ram out for better cooling and am wondering if NF3 does too. thanks.
If you're referring to the Neo2 board, I am running one module on top of the other and DDR Booster in the top slot. I had issues with spacing and heat, so I removed the heatspreaders almost immediately.

I don't have the DFI board yet. It will ship tomorrow and we'll take it from there. ;)

[xenolith]

I'd like to know how the vdimm is implemented too. I.e., is it like the NF2, NF3, or NF4 series.

Conrad, it would be really nice if you could put an X2 in it, but hey, you have a family to think about. :nono: ;)

[conrad.maranan]

Conrad, it would be really nice if you could put an X2 in it, but hey, you have a family to think about. :nono: ;)
I'll draft a parts requisition form, have all of you guys sign it, and then we'll fax it over to AMD in the morning. :lol:

[Zebo]

Here's what I'm talking about Conrad (sorry for the blurry)

http://www.csc.calpoly.edu/~smanning/mem.jpg

This can't cool at all so I run lower speeds or get mem errors due to high heats between sticks.:( I'd like them spaced better ...maybe DFI offers this?

[conrad.maranan]

That looks pretty, ummm, congested. Lose the heatspreaders and mount a small fan over the RAM. It was good for an additional 6MHz (if my memory serves me correctly) with my VX.

[Absolute_0]

Here's what I'm talking about Conrad (sorry for the blurry)

http://www.csc.calpoly.edu/~smanning/mem.jpg

This can't cool at all so I run lower speeds or get mem errors due to high heats between sticks.:( I'd like them spaced better ...maybe DFI offers this?

I like them spaced right by each other, that way my XP90 hangs over both of them and the 92mm Panaflo can cool both.

[Zebo]

Yeah I don't do fans.. My whole system only has three Nexus at 7 volts on a BIP-III. Don't like noises. Also I don't know how to remove spreaders w/oi damageing them.. they look on real permant like (crucial ballistix)

[El Snorro]

Here in holland it is already @stock so i'm gonna buy it this week, should be fun with some bh-5 (have several kits), booster, sandy, x800xt pe and a mach2 :)

[Absolute_0]

Here in holland it is already @stock so i'm gonna buy it this week, should be fun with some bh-5 (have several kits), booster, sandy, x800xt pe and a mach2 :)

sounds like a party :D

[kiwi]

Test Bed #1
DFI LANParty UT nF3 Ultra-D
AMD Athlon 64 3500+ (Newcastle)
Sapphire Radeon X850XT Platinum Edition
OCZ EL DDR PC-3200 Gold VX (2x512MB)
OCZ PowerStream 600W Power Supply Unit
Western Digital Raptor 74GB SATA HDD

Test Bed #2
MSI K8N Neo2 Platinum
AMD Athlon 64 3500+ (Newcastle)
Sapphire Radeon X850XT Platinum Edition
OCZ EL DDR PC-3200 Gold VX (2x512MB)
OCZ PowerStream 600W Power Supply Unit
Western Digital Raptor 74GB SATA HDD

How does that sound? ;)


And Test Bed #3 DFI NF4 to see maybe it is worth going back AGP lol

[Hirogen]

what's the max. vdimm of the new dfi?


is the ocz booster compatbile with it?


H.

[Vassili]

I'll hope that they locked the s-ata controller this time... that would be nice.

Official DFI SITE LANPARTY UT nF3 ULTRA-D (http://www.dfi.com.tw/Product/xx_product_spec_details_r_us.jsp?PRODUCT_ID=3160&CATEGORY_TYPE=LP&SITE=US)

I found this is on the home-page of DFI:
DFI Re-Launches The INFINITY Line To Offer High-Quality, Affordable Motherboards To Overclockers and Enthusiasts
So maybe a cheap DFI nForce3/4 Infinity? :)

[Cyprio]

So maybe a cheap DFI nForce3/4 Infinity? :)

DFI is excited to announce today that it will re-launch its INFINITY line, bringing to the market state-of-the-art motherboards featuring high-quality components and friendly overclocking features, all at a more affordable price.

The INFINITY series aims at power users on a budget. Leveraging DFI's expertise in motherboard design and the powerful performance of the popular LANPARTY series, the INFINITY line is made for the most discerning overclockers and hardware enthusiasts who demand high quality and performance without the heavy price tag.

Two INFINITY models will be introduced in the first launch: the nF4X INFINITY, a 754 pin AMD platform with PCI-e support, and nF4 INFINITY, a 939 pin AMD platform with PCI-e support. Both boards utilize the hottest nForce4 solutions from NVIDIA. A slew of models are in the process to furnish this new line soon.

“The INFINITY series offers price-conscious consumers high-end motherboards with quality components,” said Y.C. Lu, CEO of DFI. “Furthermore, its design and overclocking power makes the INFINITY boards ideal for amateurs as well as power gamers and enthusiasts.”

:)

[celemine1Gig]

...
Two INFINITY models will be introduced in the first launch: the nF4X INFINITY, a 754 pin AMD platform with PCI-e support...

That's great news! Now I know why I'm still waiting, instead of having already bought a new board for my socket754 mobile and my PCI-E video card. :)

[CrimeDog]

also spotted here:
http://www.mwave.com/mwave/skusearch.hmx?SCriteria=BA30085&CartID=done&nextloc=

[El Snorro]

I have ordered the board today, will be here tomorrow :)
Hopefully some sealstring wil arive here tomorrow so I can mount my prom on it directly and lets see if this baby can rock 'n roll :banana:

[DriveEuro]

also spotted here:
http://www.mwave.com/mwave/skusearch.hmx?SCriteria=BA30085&CartID=done&nextloc=
I called up mwave at 10:30am CST and they had 30 in stock. I ordered mine Fedex overnight for $136.66 shipped.

I think these will sell out really really fast though. People have been waiting for an NF3 939 board that is better than the Neo2. This is it.

[nweibley]

Im waiting to see that it is in fact better than the Neo2.

Alot of people got the PCI-E itch and upgraded already... so I'm thinking there will still be a few DFI boards to snatch up once I hear how they perform.

[xenolith]

IMHO, $125 is a pretty good intro price, as initial demand should be high. But if anyone is on a budget or a cheapo like me, it should come down quickly.

[xenolith]

what's the max. vdimm of the new dfi?

Looks like it's 3.1v, per screenie in another thread. No 5v jumper. :(

I see a tictac beta in the future though. :stick:

[tomati]

Would you guys be interested in a review of the DFI board? If yes, is there anything in particular that you would like to see tested?


Hi,

Could you look about the ability for fitting an huge heatsink like thermaltake big typhoon and sonic tower (~ 12cm²),

tell us what's the lowest vcore setting available,

thanks

I just ordered mine.

[pha|anx]

Test Bed #1
DFI LANParty UT nF3 Ultra-D
AMD Athlon 64 3500+ (Newcastle)
Sapphire Radeon X850XT Platinum Edition
OCZ EL DDR PC-3200 Gold VX (2x512MB)
OCZ PowerStream 600W Power Supply Unit
Western Digital Raptor 74GB SATA HDD

Test Bed #2
MSI K8N Neo2 Platinum
AMD Athlon 64 3500+ (Newcastle)
Sapphire Radeon X850XT Platinum Edition
OCZ EL DDR PC-3200 Gold VX (2x512MB)
OCZ PowerStream 600W Power Supply Unit
Western Digital Raptor 74GB SATA HDD

How does that sound? ;)


that would be AWESOME! :hump:

[DriveEuro]

I have my DFI NF3 Ultra-D up and running now. Its a bit picky about my OCZ VX ram. Refuses to boot with it in slots 2 and 4. Vdimm claims 3.2v in the bios, but it actually puts out just 3.1v (which is also read from the bios)

[sang]

I have my DFI NF3 Ultra-D up and running now. Its a bit picky about my OCZ VX ram. Refuses to boot with it in slots 2 and 4. Vdimm claims 3.2v in the bios, but it actually puts out just 3.1v (which is also read from the bios)

tell us the good/bad things when you compare this board with your Neo2.
Thanks.

[conrad.maranan]

Have you plugged in a DDR Booster yet?

[Nubius]

Mine will be here tomorrow :D

Go to www.dfi-street.com in the AMD64 motherboard section....Angry Games (for those who dont know) is a DFI rep and he's thoroughly testing his out along with a couple people who already have them.....the latest word is that only 2 SATA ports are locked...the ones nearest the CPU and boy are they like RIGHT on the CPU too.

EDIT: Ahh yes, also worth mentioning a couple people are having problems with OCZ VX line it seems....one guy says it'll ONLY work in slots 1&3 and 2&4 just give him troubles.....don't believe anyone has tested it with the booster yet

[DriveEuro]

Have you plugged in a DDR Booster yet?
Booster is a tight fit. The DDR channels are grouped together, they do not alternate like on the Neo2. This makes using the booster a pain in the ass. There is a cap that gets in the way of the booster heatsink, I was able to bend it out of the way, but it really isn't that safe since the heatsink of the booster touches it.

So far this board is a pain in the ass. It is not being consistent with my OCZ VX ram. Sometimes it will boot with it. Others, it won't. It boots fine with my OCZ Value VX though. A few times it would not boot with my OCZ VX in 2 & 4.

I really don't get this board... Its a bit aggrivating.

There are no .5 increments at all for the hypertransport multiplier. It does have voltage options for the chipset though, so I guess that cancels out the lack of .5 multipliers.

I don't like that the post screen doesn't show you the CPU speed it is booting up with. There were a few times that it booted up with my mem speed at 200mhz when I had set it in the bios at 240mhz. I guess this is more of a problem because it isn't being consistent with my OCZ VX.

I'll tinker with this a bit more tonight. Right now I'm memtesting 240mhz at 3.2v (no booster) on my OCZ VX. Hopefully it will be stable enough to sit for the next 18 hours or so, since I have work tomorrow.

[DriveEuro]

EDIT: Ahh yes, also worth mentioning a couple people are having problems with OCZ VX line it seems....one guy says it'll ONLY work in slots 1&3 and 2&4 just give him troubles.....don't believe anyone has tested it with the booster yet
That was me.

[babyelf]

Tim tested with the booster.. up to 3.5v but he hasn't max it out yet..

[DriveEuro]

Yes, but I'm not getting my OCZ VX to work in slots 2 & 4 with or without the booster. What slots was TIM using when he used the booster?

[Nubius]

lol, tim, babyelf, driveeuro.....it's like being at dfi-street all over again :p

You see the latest problems seem to be a little bit of undervolting and CPU's not working, or rather not booting into windows with HTT at 5x and CPU at default 200 because of the overall HTT hitting above 2000.....hmm....I'll be playing with mine tomorrow hopefully, but I sure hope TicTac or someone equally good can rework this BIOS....basically that's what it's coming down to.....also...I would remove my heatspreaders off my RAM if I could...last time I tried it took 30 minutes to remove 1 clip of 1 heatspreader so I said screw it and gave up...those things are on there pretty damn good.

[pha|anx]

guys, keep us updated on your experiences. a lot of people have been waiting in line a good long time for these boards, and want to see everything there is on them before buying... insecurity is called for, as nothing is infaliable... especially when it comes to computer related components lol

:rocker:

[conrad.maranan]

I'll do my best to put a full review up before next weekend. As long as I receive my board tomorrow or Monday, I should have enough time to get down and dirty with it. ;)

[terror_adagio]

A lot of the small issues, even if you want to call them that are nothing DFI cannot fix with a little BIOS update.

No mobo comes out perfect. Overall, I'm extremely happy.

[DriveEuro]

I have to use 1.8v in the bios to get 1.55v under full load.

Anyone else's board undervolting that much?

[conrad.maranan]

Maybe the board isn't undervolting. Perhaps the BIOS was just programmed with the incorrect values for voltages. In my opinion, the measured voltages from the board will always be the bottom line. BIOS can either be dead-center or slightly off target. In your case, it seems to be way off target.

[DriveEuro]

Maybe the board isn't undervolting. Perhaps the BIOS was just programmed with the incorrect values for voltages. In my opinion, the measured voltages from the board will always be the bottom line. BIOS can either be dead-center or slightly off target. In your case, it seems to be way off target.
Yeah, I've been thinking that. No one else at DFI Streets is having this problem.


EDIT!

IMPORTANT UPDATE!!!!

http://www.dfi-street.com/forum/showpost.php?p=137919&postcount=507

Vcores on this board are messed up!

1.550 +

0.1 = 1.47v
0.2 = 1.56v
0.3 = 1.61v
0.4 = 1.72v

[terror_adagio]

Nothing a new BIOS won't fix.

[DriveEuro]

Nothing a new BIOS won't fix.
Right, but assume someone buys the board and sets vcore to like 1.5v and their cpu still doesn't boot. This could help out someone. I couldn't get my rig running (overclocked) till I pumped 1.8v into the cpu.

[Nubius]

This is a begging plea to TicTac or anyone equally as good :p We need a new BIOS....all our respect are belong to you :D

[DriveEuro]

What would we need a new bios for aside from Vcore?

I had a problem last night where the computer kept booting up with DDR200 instead of DDR400. i dunno what's up with that. But I think that there should be a multipler setting for 200, cuz it doesn't seem to always lock at 200.

[terror_adagio]

What would we need a new bios for aside from Vcore?

I had a problem last night where the computer kept booting up with DDR200 instead of DDR400. i dunno what's up with that. But I think that there should be a multipler setting for 200, cuz it doesn't seem to always lock at 200.

Let DFI tech know of any issues you have.

[DriveEuro]

Crap. Just tried to cold boot. It won't boot. It just turns on and makes that clicking noise on my speakers. :(

[Nubius]

What would we need a new bios for aside from Vcore? Because that doesn't seem to be the only current issue after reading numerous complaints over at DFI-street....seems right now the BIOS just needs a little tweakin, not only for voltage readings, but for stability too.....plus the RAM thing is weird being that a lot of people can't seem to be able to blatently set 200, but beyond that it seems a lot of people are having trouble with running 1T.......BIOS just needs to optimized a tad is what I'm saying

After the troubles you've had and the posts I've seen you made, you really had to ask why I would say a new BIOS is needed?

[NEC]

Test Bed #1
DFI LANParty UT nF3 Ultra-D
AMD Athlon 64 3500+ (Newcastle)
Sapphire Radeon X850XT Platinum Edition
OCZ EL DDR PC-3200 Gold VX (2x512MB)
OCZ PowerStream 600W Power Supply Unit
Western Digital Raptor 74GB SATA HDD

Test Bed #2
MSI K8N Neo2 Platinum
AMD Athlon 64 3500+ (Newcastle)
Sapphire Radeon X850XT Platinum Edition
OCZ EL DDR PC-3200 Gold VX (2x512MB)
OCZ PowerStream 600W Power Supply Unit
Western Digital Raptor 74GB SATA HDD

How does that sound? ;)


Good :up:

[terror_adagio]

What nForce drivers is everyone using with this board?

Which are the best: 5.10 Offical or 6.37 Beta.

[BobyJo]

www.MWave.com

Has them in stock at this time. I ordered one last night, so there are some in stock.

[conrad.maranan]

After extensive testing with this board, I will be holding off on my review until a new revision to the BIOS is released. My opinions would be somewhat harsh if based on the BIOS shipped with the board.

Since I have found ways of getting around a number of issues, please feel free to request for assistance via PM if you are having problems. I'll see if my fixes can get you up and running. :)

My conclusive data to date:

PROS
- A net gain of 60MHz (CPU) and 6MHz (RAM) over the MSI K8N Neo2 Platinum

CONS
- BIOS resets TRP to 3, regardless of value entered. TRP needs to be re-adjusted using A64Tweaker.
- Using DIMM slots 1 and 2 equates to single-channel memory output. Although CPU-Z will recognize this setting as dual-channel 1:1, SuperPI 32M performance will dictate otherwise - and believe me, over 45 minutes to run 32M at 2560MHz 2-2-2-10-1T-1:1 is a sad shame.
- BIOS does not include 200MHz (1:1) option regardless of DIMM slot combination used.
- Frequencies set in BIOS tend to adjust to higher clocks after 250MHz once in Windows. Clocks will have to be reset with ClockGen to the desired frequency.
- Voltages for CPU are about a mile off. Using the MSI K8N Neo2 Platinum, my CPU would require 1.550V + 10% in order to run 2600MHz with stability. 1.550V + 13% would result with BSOD. With the DFI board, the same CPU needs to be set at 1.550V + 0.4V in order to run 2600MHz and up with stability. 1.550V + 0.4V = 1.950V in BIOS. CPU-Z will show a voltage reading of 1.780-ish with the DFI using the aforementioned BIOS settings. A reading of anything over 1.675-ish in CPU-Z with the K8N Neo2 Platinum would cause the system to crash when attempting to run SuperPi 1M.

All of the quirks are BIOS-related. The board does appear to have a lot of potential, though. If you are planning to purchase this board, I suggest that you hold off until the next BIOS release.

[Absolute_0]

Ram slots 1 and 2 give you single channel ram? Uh oh, that's what i'm using!

DOH

No wonder my 1Ms suck so hard LOL
:slapass: at CPUZ for telling me it was dual channel!

[conrad.maranan]

Yeah. I figured it out the long way: running SuperPI 32M 3 times at different clocks. :lol:

The manual specifically states that DIMM slots 1 and 2 are for single channel. But the information it reports to CPU-Z and ClockGen is misleading. SiSoft Sandra will read it correctly, though.

[Absolute_0]

Conrad, i just tried putting my ram in dimms 1 and 3 instead of 1 and 2. Well, at my current settings (2.86 Ghz 1.56 volts, which are pretty stable) it was compltely unstable, SP1M couldn't run and prime95 failed instantly.

I tried dimms 1 and 4 and the computer beeped at me a bunch...

So i'm back to 1 and 2.
Which dimm slots are you using?

EDIT:
Call me slow but i'm figuring out that 1 and 3 was the correct combination and running dual channel made my current settings unstable; i'll have to tweak new ram timings and stuff... right?

[conrad.maranan]

The reason why DIMM slots 1 and 3 cause your system to crap out is because the BIOS sets itself to DDR200. Look at the BIOS screen the next time you boot up and see if it reads DDR200. With my VX in 1 and 3 - DDR200. With some TCCD in 1 and 3 - DDR200.

The BIOS needs a major overhaul as far as memory settings and voltages are concerned. Let's see if anything comes out in the next week. :)

EDIT:
So i'm back to 1 and 2.
Which dimm slots are you using?
No slots. I packed the board back in the box for now.

[Absolute_0]

No, it is detecting as DDR333, because i'm running the 166 divider

I've scaled back to 2.7 Ghz so at least i'm stable now

Conrad, as i'm booting up, it says 333 and SINGLE CHANNEL.... and i just got a 35 second 1M at 2.7 Ghz, i'm thinking something is wrong here. This is in dimm slots 1 and 3, 300x9, 166 divider, sloppy timings on.

OK, well i just put the ram back in 1 and 2 and EXACT same bios settings, i did SP1M and got 32.64

So these are my results:
9x300, 166 divider, sloppy ram timings and 2T
dimm slots 1/3 = 34.64 SP1M
dimm slots 1/2 = 32.64 SP1M

i'm sticking to 1 and 2

[Judaeus Apella]

Hey everyone! I was hibernating until this mobo came out, but now I’m back and ready for action! This is the system I intend to build:

- DFI Lanparty UT nF3 Ultra-D Motherboard
- AMD Athlon 64 FX55 San Diego 1GHz FSB Socket 939 Processor
- OCZ EL Platinum Revision 2 Dual Channel Kit 184-Pin 1GB(512MBx) DDR PC-3200
- Thermalright XP-120 Heatsink
- SUNON KD1212PMB1-6A 120mm Ball Case Cooling Fan - OEM
- Ryanpower2 550W ActivePFC - Black CableFREE™ Power Supply, ACR-PS2100
- nVidia PNY Verto GeForce 6800 128mb AGP Graphics Card

Someone told me that this motherboard has some issues with memory. Anyone know if the memory above will work with it? Or will there be a problem. I also read a new BIOS is being made by the company to fix the problem but it will take a little bit of time before its released. If there is an issue with this memory, does anyone know when that BIOS is coming out?

Also, is this the best memory I can get for gaming? I've heard more good things about OZC than Corsair or any other brand, especially Samsung chip based RAM, which is what the memory is that I'm getting. I plan on buying a second set of them in a month when I can better afford it.

[Giasan]

Second run of the "have a million problem board":

http://img238.echo.cx/img238/7444/3d23004sr.jpg (http://www.imageshack.us)

http://img238.echo.cx/img238/9281/atweak3000re.jpg (http://www.imageshack.us)

HTT: 3x (over 1000mhz there is a wall)
Chipset @1.75 volts
Mem 2x512 OCZ platinum PC3200 REV 2 TCCD @2T (1T is a problem for a lot of memory over 230mhz). The 2T seems not hurt board performance.
Problem whit TRp, if set to 2 the board set 3, if set 4 the board set 5, if set 3 seems that the board set 3.

Problem of cold boot. If at first power up the board wan't boot up and all led blink, simply wait 3 second and just push reset bootton whithout power down the PC.

Same color = dual ch. (orange orange - yellow yellow)

The bios is very far from the NF3 250gb 754, there isn't many option.

This bios need a good revision, please OSKAR TICTAC JESS help us.

Bye.

Exuse conrad for the duplicate tread.

[Jacke2k]

Do eny one you knowe when this board comes to sweden ?

[split]

Do eny one you knowe when this board comes to sweden ?

This (http://www.trend4pc.de/index.php) place ship?

I see someone in the UK got one shipped from here (http://www.icomputers.nl/product_detail.asp?A_ID=2364&S_ID=411&SS_ID=888&SSS_ID=1091).

This bios need a good revision, please OSKAR TICTAC JESS help us.

Thinking...someone might mod it if they had it.
All that's on the DFI site atm is the beta one, so might be a good idea to upload the original.

I can host it if you cant.

[kryptobs2000]

Well, I'm gonna get this mobo, but does it seem like it's strickly bios issues or hardware issues? Ala... can it probably be fixed with a bios update, or is there going to be a new board revision I should hold out for?

[conrad.maranan]

I have not noticed anything wrong at the hardware level. The board basically has the same feel as its 754 counterpart. There isn't a 24-pin PSU requirement or a 4.0VDIMM jumper to worry about. Mounting a huge heatsink isn't a problem and will leave ample room around the two SATA ports that neighbor the CPU socket. When installing to a phase change unit, I had to install the SATA cable first in order not to experience any difficulty after the evap block and compressed insulation were in place.

I doubt DFI will require a second revision for this board. They just need to put out a good BIOS and it's a done deal.

[conrad.maranan]

Okay. I've been playing with the new 6/28 BIOS since 90 minutes ago and it looks to be very promising. The changes I have noticed as of this writing are as follows:

- TRP remains at the value set in BIOS and no longer reverts to a value of 3
- VCORE tuning (+0.1, +0.2, +0.3, etc.) is still inaccurate but seems to have slightly improved
- Clocks/Frequencies stay true even beyond 250MHz - no need to re-adjust with CPU-Z
- 200MHz/DDR400 RAM frequency is now a selectable option in BIOS
- DIMM slots 1 and 2 are now officially supported as 200MHz/DDR400
- 4°C lower CPU temperatures at 2600MHz (10x260MHz) measured with UEI DT200 and K-type probe

If the 6/28 BIOS is an indication of anything, this board is going to shine with the BIOS revisions that follow. :D

EDIT: The Quick Power On Self Test option in BIOS is somewhat buggy. What you want to do is Enable this option so that the boot up time is faster. However, even after selecting Enable, it will revert back to Disable. This is just a small quirk and I'm sure it will be an issue addressed in the next release. Just press ESC once the RAM test starts rolling and it'll skip the test.

[Giasan]

Okay. I've been playing with the new 6/28 BIOS since 90 minutes ago and it looks to be very promising. The changes I have noticed as of this writing are as follows:

- TRP remains at the value set in BIOS and no longer reverts to a value of 3
- VCORE tuning (+0.1, +0.2, +0.3, etc.) is still inaccurate but seems to have slightly improved
- Clocks/Frequencies stay true even beyond 250MHz - no need to re-adjust with CPU-Z
- 200MHz/DDR400 RAM frequency is now a selectable option in BIOS
- DIMM slots 1 and 2 are now officially supported as 200MHz/DDR400
- 4°C lower CPU temperatures at 2600MHz (10x260MHz) measured with UEI DT200 and K-type probe

If the 6/28 BIOS is an indication of anything, this board is going to shine with the BIOS revisions that follow. :D

EDIT: The Quick Power On Self Test option in BIOS is somewhat buggy. What you want to do is Enable this option so that the boot up time is faster. However, even after selecting Enable, it will revert back to Disable. This is just a small quirk and I'm sure it will be an issue addressed in the next release. Just press ESC once the RAM test starts rolling and it'll skip the test.



Next round cmon:

http://img154.echo.cx/img154/3360/325mhzpi2fg.jpg (http://www.imageshack.us)


http://img154.echo.cx/img154/5845/3dmark013261sm.jpg (http://www.imageshack.us)


Original bios the other beta better stay at DFI site. With 0627 i can't boot whit raid set and whit eide disk, only black screen after post.

Whit 6/28 if i set 250, mobo after power up in safe mode and i enter the bios have set 353mhz on cpu :confused: :confused: .

Virus waring can't put on disable
Quick power can't put on enable

So i return on the original and i'm wait for better days.

Bye

[split]

Some kind person has stuck a BIOS Page (http://www.lejabeach.com/DFI/dfinf3s939/dfnf3bios.html) up in case you need to flash back.

[Giasan]

Some kind person has stuck a BIOS Page (http://www.lejabeach.com/DFI/dfinf3s939/dfnf3bios.html) up in case you need to flash back.


Good work, but be aware from 0627 (not 627) and never see 628 that is the worst.

Bye.

[conrad.maranan]

6/28 is a drastic improvement over 6/02 and 6/27.

[Nubius]

6/28 is a drastic improvement over 6/02 and 6/27. DFI reps Angry_Games and RGone 100% completely disagree with that....they were trying to get the 628 BIOS off the website before anyone downloaded it because it was apparently causing some computers to boot up in a high HTT like 340 even if it was set much lower.....the words "Bios corruption" were also slung around a bit....as far as I've been following at the official NF3 page at dfi-street the 628 BIOS is the worst of them all.

[Nubius]

I just tried the latest 629 BIOS for this NF3 board......it was way more flakey than the original BIOS for me.

The CPU Voltage was reading at 1.31v on default instead of 1.28 so it seems like that's slowly getting straightened out..

My RAM defaulted at 2-3-7-5-2T timings believe it or not.....and upon trying to change them to their normal 2-2-2-5 stock timings at stock speeds the computer would refuse to post....seems like any little touch I tried to do to it, and it's refuse to post so I just flashed back to original BIOS since I made sure to make a copy of it on a floppy before I flashed with this new one.

[conrad.maranan]

Hmmm. My understanding was that the 6/27 BIOS was the one that would likely cause BIOS corruption. I couldn't pass 230MHz without looping the nVIDIA RAID screen with 6/27. I doubt my case is special, but I'll read up on what has been said at DFI-Street later on.
I just tried the latest 629 BIOS for this NF3 board
There's a 6/29 BIOS that was released today?

[terror_adagio]

All BIOSs that have been released thus far are BETA. So don't expect them to be stable people.

[conrad.maranan]

I am up to speed with the latest happenings at DFI-Street. I figured I might as well register an account so I can defend my opinion regarding the 6/28 BIOS. 6/29 is now saved onto a floppy, but I doubt I'll be up for playing with it tonight. We'll see. :)

[pha|anx]

thanks for all the info and feedback conrad.maranan!! you have no idea how much this means to everyone!

:thumbsup:

[Fission]

I am still loving the orignial BIOS, since any of the issues seem to have sufficient work-arounds. Lots of people at the DFI forums are screaming and freaking aout about how this board is a lemon, but whatever. I did 340HTT prime stable the 3rd day it was on the market, I am happy with that as a good release. BIOSes are important sure, but clockgen and A64 tweaker are better ways of overclocking, and the board is at least stable at high overclocks, and delivers nice stable voltages.

I am waiting for the bios that give it a 10:9 divider to come out, or if there is one that comes beating down my door with a 5% speed boost guarantee, maybe then I'll change it.

Mixed reviews on the 628, don't know who to trust... I think I'll let it gain some maturity before I go looking for a new BIOS just yet, totally happy with the current BIOS.

Accurate temps would be nice, but will that ever happen on an nForce, or any motherboard for that matter?

[conrad.maranan]

Fission - Can you do me a favor and run SuperPI 32M and report back with your CPU speed (MULTI x HTT), RAM speed (dividers and timings, too), and your 32M time.

Thanks. :up:

[eyepiz]

Hmmm. My understanding was that the 6/27 BIOS was the one that would likely cause BIOS corruption. I couldn't pass 230MHz without looping the nVIDIA RAID screen with 6/27. I doubt my case is special, but I'll read up on what has been said at DFI-Street later on.

There's a 6/29 BIOS that was released today?

what do you meen by "without looping the nVIDIA RAID screen"

I'm having touble booting up on power on... safe mode settings kick in but only on power on and not restart.

Strange

[conrad.maranan]

The nVIDIA RAID screen that immediately follows the initial boot-up screen would go back to the initial boot-up screen...and then to the nVIDIA RAID screen...and then to the initial boot-up screen...and then to the nVIDIA RAID screen...

A continuos loop of the above is all I got with 6/27. If it happens to you in 6/02 or 6/28 at startup but not at restart, try this:

- Turn off the computer, but not the power supply
- Remove one module of RAM (from the DIMM slot farther away from the CPU)
- Turn on computer
- Shutdown computer
- Re-install the memory module that was removed
- Turn on computer

The above should fix the problem. ;)

[eyepiz]

thanks,

I will give that a shot maybe it fixes my powering on in safe setting issue.

[conrad.maranan]

What BIOS are you using?

[eyepiz]

What BIOS are you using?

Right now the original but I have tried them all 27,28 & 29 same thing with the safe setting loaded at power up.

Strange cause I'm 100% stable passed 6hrs memtest no errors, superpi 32M, OCCT & 3Dmark05 and most important BF2 played 4hrs straight with no problems and we all know how much of a systems how BF2 is.

all test @ 9 X 290HTT 3XLTD 1:1 @SPD 2.5-4-4-10 2T

[conrad.maranan]

So, did it work?

[eyepiz]

:( no same thing still

[G H Z]

I've got one of these boards but it won't POST, just continually beeps. I think its a power issue somewhere. I tried dropping voltages (OCZ 520) they were cranked, but that didn't help.

It's interesting your results with the 6/28 BIOS after reading the DFI forum thread last night. To be honest I trust your results over A_G's. RGone I'm not sure about. Does anyone know any more details or exactly what was happening that made them want to pull it off the DL? I know it was a high HTT set during boot but was there any other info?

Do you flash using multiple switches like those guys Conrad?

[eyepiz]

G H Z

Try booting with one stick of ram in the dimm closest to the cpu.

I had that issue when i first hooked this mobo up.

[conrad.maranan]

Try my solution in Post #90 - that solved my beeping problems, too. And speaking of beeps, I'm probably going to end up yanking out that piezo buzzer real soon. That thing is annoyingly loud.

As far as 6/28 is concerned, I didn't have any of the problems that was mentioned over at DFI-Street. As soon as I flashed over, I went straight into BIOS and replaced all of the AUTO values for my own combination of settings. The only thing that truly bugged me was the Quick Power On Self Test issue. Other than that, everything worked well for me.

As far as flashing is concerned, I usually use WinFlash. But for this board, I've been using the Diamond Flash installer app that's in their download section.

[Fission]

Fission - Can you do me a favor and run SuperPI 32M and report back with your CPU speed (MULTI x HTT), RAM speed (dividers and timings, too), and your 32M time.

Thanks. :up:

In any particular configuration?

[conrad.maranan]

Any configuration is fine. :) 2x512MB of RAM preferably.

[G H Z]

G H Z

Try booting with one stick of ram in the dimm closest to the cpu.

I had that issue when i first hooked this mobo up.

That did the trick, guess I lost interest very quickly after reading about all the problems this board has off the shelf. I was expecting another Lanparty NF3 experience, but not this time.

Conrad YHPM

[conrad.maranan]

Don't lose hope just yet. It's too early to draw conclusions.

[Fission]

Conrad, what Nforce drivers are you using?

[conrad.maranan]

Version 4.27 for Windows 2000.

[Fission]

At your request m'lord:
http://img158.imageshack.us/img158/8083/32m3ar.th.jpg (http://img158.imageshack.us/img158/8083/32m3ar.jpg)

[conrad.maranan]

Amazing. And that's with the 6/02 BIOS? Regardless of what type of RAM I used or what slots they were placed in, I couldn't pull off anything better than 45 minutes at 10x260MHz. Usually, I can pull it off in 28 minutes or less.

Strange stuff.

[Giasan]

I've got one of these boards but it won't POST, just continually beeps. I think its a power issue somewhere. I tried dropping voltages (OCZ 520) they were cranked, but that didn't help.

It's interesting your results with the 6/28 BIOS after reading the DFI forum thread last night. To be honest I trust your results over A_G's. RGone I'm not sure about. Does anyone know any more details or exactly what was happening that made them want to pull it off the DL? I know it was a high HTT set during boot but was there any other info?

Do you flash using multiple switches like those guys Conrad?


Whit the 0627 (not the 627, it is another) boot up but raid and ide disk don't work.
Whit the 628 if i disable virus warning in the bios and enable quick boot, save and exit at restart this think return in original state (virus warning enable , quick boot disable). Othr whit 628 if i set 250 of HTT i can't boot up. When the system restart i safe mode and i can enter the bios i see that the mobo set 353 of htt.

Bye

[conrad.maranan]

I flashed from 6/28 back to 6/02, and then back again to 6/28 to see if the 353MHz HTT issue would manifest itself with my board. I shutdown my computer and power supply, powered up my system, went into BIOS to change CPU multipliers and HTT, and then started up Windows. After Windows loaded up, I would open up ClockGen and CPU-Z to verify that my clocks were exactly as I had set them in BIOS.

- 11x200MHz: clocks remained exactly as set in BIOS
- 11x210MHz: clocks remained exactly as set in BIOS
- 11x220MHz: clocks remained exactly as set in BIOS
- 11x230MHz: clocks remained exactly as set in BIOS
- 11x240MHz: clocks remained exactly as set in BIOS
- 10x250MHz: clocks remained exactly as set in BIOS
- 10x260MHz: clocks remained exactly as set in BIOS

I have yet to experience the 353MHz HTT bug with the 6/28 BIOS. Indeed, this is strange and my board seems to be exempt from the dilemma.

UPDATE:
I flashed over to the 6/29 BIOS and noticed the following:

PROS
- Quick Power On Self Test issue has been resolved
- Virus Warning has been resolved
- 125MB/s SiSoft Sandra Memory Bandwidth increase over 06/28 BIOS
- SuperPI 1M stability at 20MHz higher clocks than that realized with 06/28 BIOS; CPU stability similar to 06/02 BIOS

CONS
- Ethernet/Network adapter fails to function; no Internet
- VCORE tuning (+0.1, +0.2, +0.3, etc.) is still inaccurate
- SiSoft Sandra Memory Bandwidth is 100MB/s less than exact same system on an MSI K8N Neo2 Platinum board with replicated frequencies and timings

So, are we going to see a 6/30 or 7/01 BIOS? I like the direction this is all heading and it does seem like progress is being made. 6/29 would be the one I'd run if I can get around the network adapter issue. I'll re-flash again tomorrow and see if my luck changes.

[Badong]

Well i was reading thru the thread. It took DFI almost a year to release their motherboard. I remember K8N Neo2 Platinum were available at August 2004.
Yet, DFI managed to release this board with seriously undeveloped/untested BIOS.

Amazing. Good luck to the owners!

[eyepiz]

That did the trick, guess I lost interest very quickly after reading about all the problems this board has off the shelf. I was expecting another Lanparty NF3 experience, but not this time.

Conrad YHPM

I think DFI Boards are made to initial boot with one stick of memory in that dimm. It took me two days to get it to post. I had finally gave up and call DFI Tech support They gave me the procedure.

Maybe thay should put that in the manual cause some of us are not psychic.

but the board rocks!! I'm at 9X 290HTT 1:1 100% stable. there's just a stupid booting issue I think it might have to do with raid-0

[Nettwerk]

Im getting my DFI nf3 mobo today and a 3000 venice cpu. Which bios do you guys recommend to use untill they fix the problems?

[eyepiz]

Try the 602 fist see how it handles. and if you have problems try 627 I have had issues with 628 & 629.

[Fission]

Amazing. And that's with the 6/02 BIOS? Regardless of what type of RAM I used or what slots they were placed in, I couldn't pull off anything better than 45 minutes at 10x260MHz. Usually, I can pull it off in 28 minutes or less.

Strange stuff.

That is the 6/02, completely stable at 10X287, 8x340, 9x312, really no problems in 1T at any settings, so long as my RAM stays under 265, and if I go over 295HTT, it is better to bump HTT voltage to 1.75. Northbridge cooler will get switched out for a modded socket A cooler within a week or so, it is too hot for my liking with that paper thing heatsink it curently has.

In 1/3 (orange ones) the RAM wouldn't top 250, which I would have thought was the RAM if I hadn't had the RAM stable at 264 in my old VIA, but in 2/4(yellow) they are good to 265 at same timings. I use 1.50 + .4 in BIOS for 1.64v core, and to go to 2.9 I go to 1.525+.4 for 1.69. It is on an XP-90, so it gets a little warm much over 1.7v, but I have stil run it prime stable 1 hour at 1.744vcore at 2.912Ghz, but I don't like 48 degrees idle, 55 load, that scares me. I have it set to shut down at 60*, but haven't had it do that yet.

I do all of my overclocking/dividers through Clockgen/A64Tweaker, only use BIOS for timings (which stay where I set them) and voltages.

I am coming off of a VIA KT800, which did alright, but was dirt compared to this thing, and I am personally satisfied with it's performance. 29s 1MB seems low for 2.87Ghz, and My Aquamark CPU score of 13,444 also seems low for teh above configuration, but it is still in it's infancy.
Those who thought it would be perfect have unrealistic expectations.

[kryptobs2000]

DFI 754 was perfect and not nearly as buggy at the initial release, but I have high expectations from this board. I say withen a couple weeks it'll be as good as it's 754 counterpart, and same with it's nf4 counterpart minus the randomly killing ram and board dying problems.

[Supertim0r]

I just ordered mine an hour ago :)
I ordered an other 3000+ & 3200+ (0517)
can't wait to test it ! :woot:

[Fission]

I just ordered mine an hour ago :)
I ordered an other 3000+ & 3200+ (0517)
can't wait to test it ! :woot:

Five bucks says he loves it.

[Absolute_0]

Fission, i've got my HTT voltage at 1.75 also. Smart guardian reports load temps no higher than 35C for my chipset, I dont know how accurate that is, but how hot is your chipset getting? Do you see upgraded chipset cooling as a must for anyone running 1.75V on the chipset?

[conrad.maranan]

I run 1.85V on the chipset with the stock heatsink/fan. I replaced the generic thermal compound with Arctic Silver. Seems alright to me. :)

[Fission]

I also run it at 1.85 on stock cooling, and it hasn't had a problem per se, yet it is new, and the heatsink on the nb feels about 80 degrees, and even with a 80cfm 80mm right over it it is still VERY hot to touch, despite BIOS always saying 27-32c. I think some epoxy and AS5 with the copper slug off of a old Volcano will please it, if the black aluminum 3mm thick postage stamp on there now can't. I will have to cut the fins on the heastink in a little staircase or pyramd because of videocard, but I think it will be better, or back on with stock. I'll take some pics before I mount it. I guess I could screw it on too.

[winston856]

DFI 754 was perfect and not nearly as buggy at the initial release, but I have high expectations from this board. I say withen a couple weeks it'll be as good as it's 754 counterpart, and same with it's nf4 counterpart minus the randomly killing ram and board dying problems.

Yes, this is true with most DFI boards. Some are alright on the stock BIOS, most take a few BIOS's before they get 'dialed in' :toast:

[Walrusbonzo]

Just ordered one of these to replace my 754 DFI NF3 board. Been looking for a good reason to upgrade my CPU as my current heatspreaderless A64 3000+ is terrible, needs 1.85v to do 2.5GHz, watercooled to 42C full load!!!

So ordered a 3000+ Venice too, fingers crossed for 2.6GHz+ without having to use silly voltages.

[winston856]

Should not be a prob Walrusbonzo, once there's a good BIOS out for this board it's going to smoke. I want to pair it with a 3700+ SD chip :bananal:

[Supertim0r]

Five bucks says he loves it.

I finally changed my mind and ordered a "random" 3200+ stepping (but I kept the 3000+ 0517 order)
:D

[winston856]

OT: Supertim0r I LOVE your steam sig :ROTF:


Back on topic.

[Supertim0r]

:p: :p:

[G H Z]

Amazing. And that's with the 6/02 BIOS? Regardless of what type of RAM I used or what slots they were placed in, I couldn't pull off anything better than 45 minutes at 10x260MHz. Usually, I can pull it off in 28 minutes or less.

Strange stuff.

Conrad have you played with it anymore? Did you run any benches on 6/02 before you flashed back and fix this apparently single channel or underspeed Ram issue?

45min @ 2.6ghz is terrible.

[split]

I see someone here (http://www.dfi-street.com/forum/showpost.php?p=149158&postcount=1241) stuck a Vantec on his northbridge.

You may to consider your vid card cooler if your using a ATI Silencer.

[G H Z]

Been tesing for the past few hours.

Setup:

SD 4000
BIOS 6/29
Mushkin LII 2x512

Pretty dissapointed in DFI and their lack of a decent BIOS on release and after three weeks we still have nothing decent.

1 - Board will POST with Dimm 1&3 populated @ 100mhz Ram frequency ONLY and also shows as Single Channel during POST. This is supposed to be Dual channel according to the Manual. Upping or setting max Ram clock to 200Mhz it will fail to boot with same memory config.

2 - Dual channel is apparently achieved using Dimm's 1&2.......so much for the manual. I don;t like Dual channel with side by side modules, too much heat and it's not as easy to dissapate. One very good reason NF4 is better.

3 - Playin around with Clockgen I noticed I could set a stable (PI 1M) speed of 237x12 but I could not boot setting anywhere near that speed via BIOS. 220x12 was all it could manage. I've never seen this big of a discrepancy between a Clockgen and BIOS overclock.

My air tests did give me the impression that the board itself can fly. But DFI better get something going pronto or this is gonna find a new home.

[G H Z]

I see someone here (http://www.dfi-street.com/forum/showpost.php?p=149158&postcount=1241) stuck a Vantec on his northbridge.

You may to consider your vid card cooler if your using a ATI Silencer.

Uh, the stock chipset cooler is way better than that thing :slapass:

[split]

Uh, the stock chipset cooler is way better than that thing :slapass:

I wouldn't have thought there was much if anything in it as the Vantec runs at 5000rpm.
If you read that thread you'll see a few people having noise problems with the stock fan (goes out of balance)

[Bandit1]

umm,guys,i hope this program isnt accurate.note it saying core 2 vcore is 1.6!!??i mean,can both core temps be monitored like that?

cpu-z doesnt tho.it shows vcore 1.20 on both core..

thing is i am geting faily high temps at default 1.2 vcore(says bios) even?yes,im keeping it bottomed out till i figure my temps for sure.so,what ya think of accuracy of bios temps?thx..stock bios,big typhoon cooler..

oh,x2 4400+

erratic changes in bios temp too.it will sometimes dip to 16c?!

http://users.adelphia.net/~ace7/images/abcdef.jpg

http://users.adelphia.net/~ace7/images/abcdefg.jpg

[Disposibleteen]

OT: Supertim0r I LOVE your steam sig :ROTF:


Back on topic.
i would like your sig winston if it would have identified me correctly, im running Opera 8.1 and it is set to identify as IE6 so it tricks your little sig there.

Sorry back on topic:

[winston856]

i would like your sig winston if it would have identified me correctly, im running Opera 8.1 and it is set to identify as IE6 so it tricks your little sig there.

Sorry back on topic:

touché

Did identify your IP, ISP and OS correctly? ;)

On topic again :D

[conrad.maranan]

Conrad have you played with it anymore? Did you run any benches on 6/02 before you flashed back and fix this apparently single channel or underspeed Ram issue?

45min @ 2.6ghz is terrible.
Sorry for the delay. I've been out of town.

Anyway, I think 6/29 is the best BIOS out of the three available releases. It turns out that my Internet connection was just down at the time when I flashed over to 6/29. Therefore, 6/29 does not kill your ethernet.

Other than SuperPI 1M and 32M, I only ran Sandra to find out where my bandwidth stood with 6/02. I'm done with 6/02 and will never go back to it again. I'm hoping to resume testing again on Tuesday. :cool:

[terror_adagio]

Sorry for the delay. I've been out of town.

Anyway, I think 6/29 is the best BIOS out of the three available releases. It turns out that my Internet connection was just down at the time when I flashed over to 6/29. Therefore, 6/29 does not kill your ethernet.

Other than SuperPI 1M and 32M, I only ran Sandra to find out where my bandwidth stood with 6/02. I'm done with 6/02 and will never go back to it again. I'm hoping to resume testing again on Tuesday. :cool:

6/29 isn't on their beta site? Where did you get it?

[kryptobs2000]

http://www.lejabeach.com/DFI/dfinf3s939/dfnf3bios.html

[Supertim0r]

Well...i've got the board and 1 cpu (waiting the 3000+ 0517)
cpu = 3200+ LBBLE 0521GPKW
I don't know if it's the cpu or the board but it's the baddest crap i've had with AMD :(
cannot boot windows @ 2.2ghz

[Supertim0r]

could someone confirm that same colored dimm = dual channel ? :confused:

[conrad.maranan]

Yes. Same colored DIMMs = Dual Channel

For some strange reason, the order of DIMM slots is setup weird: DIMM 1, DIMM 3, DIMM 2, DIMM 4. So, DIMM 1 + DIMM 3 = Dual Channel and DIMM 2 + DIMM 4 = Dual Channel. The manual is misleading as well.

[Supertim0r]

Yes. Same colored DIMMs = Dual Channel

For some strange reason, the order of DIMM slots is setup weird: DIMM 1, DIMM 3, DIMM 2, DIMM 4. So, DIMM 1 + DIMM 3 = Dual Channel and DIMM 2 + DIMM 4 = Dual Channel. The manual is misleading as well.

any tweak/settings to share with me :(
I've got big problem with this board and a venice 3200+
I'm using the 629beta
Should I put the original bios ?

[conrad.maranan]

What do you need to know? The 6/02 BIOS is garbage in my opinion. 6/29 is where it's at. Just replace as many of the AUTO settings with a value that you know will work. :)

[Supertim0r]

What do you need to know? The 6/02 BIOS is garbage in my opinion. 6/29 is where it's at. Just replace as many of the AUTO settings with a value that you know will work. :)

well...I got a corrupted hdd now...cannot overclock with the bios, freeze half the time with clockgen, voltage @ 1.62(bios) 1.58 windows...
It's more simple to ask the problem I haven't got :(

[Supertim0r]

I think i'll got pci-e...
Is there as much problem as the nf3 with the nf4 ultra-d ? :confused:
Maybe an A8N SLI Premium :|

[G H Z]

DFI NF4's are very good, the best actually aside from the occasional high vdimm issue.

[LittleDevil]

Does anyone have pictures of bios settings? Please post it.

Does anyone have redline and DFI NF3?

LP

[kryptobs2000]

So far I'm having pretty good experiences with the board. I'm running a 3200+ venice at 2750mhz stable, tho I'm running a 133 divider with my bh5, I got to about 230mhz with it set to 3.2v so thats not too bad, everything was on auto, I THINK it was set to 1T manually but I forget. I'm gonna do the 3.3v = vdimm mod sometime today probably and then I'll report back. One issue I'm having though and I think it's the mobo's fault. The processor is also new but I'm not about to try it in my craptacular Neo2. But I'm 100% stable at 2750mhz, 2800mhz is almost 100%, but at 2850mhz I bsod almost immediatly. I can't believe that that the processor just all of a sudden becomes so unstable, and I'm hoping thats not the case and that there's more in it. I'll do some tweaking and let you guys know. I'm using the 629 bios btw.

edit: I do have issues with 1T just verified, bsod when loading windows when running 230mhz with 1T set, 2T is fine. 1T to 2T has never gained me anything on my bh5 either so it's gotta be something with the board.

[G H Z]

Does anyone have pictures of bios settings? Please post it.

Does anyone have redline and DFI NF3?

LP

Lots of problems with Redline on this board according to posters in DFI forum.

[Fission]

So far I'm having pretty good experiences with the board. I'm running a 3200+ venice at 2750mhz stable, tho I'm running a 133 divider with my bh5, I got to about 230mhz with it set to 3.2v so thats not too bad, everything was on auto, I THINK it was set to 1T manually but I forget. I'm gonna do the 3.3v = vdimm mod sometime today probably and then I'll report back. One issue I'm having though and I think it's the mobo's fault. The processor is also new but I'm not about to try it in my craptacular Neo2. But I'm 100% stable at 2750mhz, 2800mhz is almost 100%, but at 2850mhz I bsod almost immediatly. I can't believe that that the processor just all of a sudden becomes so unstable, and I'm hoping thats not the case and that there's more in it. I'll do some tweaking and let you guys know. I'm using the 629 bios btw.

edit: I do have issues with 1T just verified, bsod when loading windows when running 230mhz with 1T set, 2T is fine. 1T to 2T has never gained me anything on my bh5 either so it's gotta be something with the board.

I had problems with 1T above 280 in slots 1/3 but switching to the yellow ones(2/4) completely fixed it up for me. It may sound strange, but watching the temps, don't be afraid to give your chip stability through voltage. It's of course always a little risky, but voltage does help stability at higher clocks, plain and simple.

[kryptobs2000]

Well now it's failing prime at 2800mhz everytime in 3 min. So that actually makes me feel better. It is the processor probably, maybe I just had it clocked differently and didn't realize it before or something, and it makes since it'd bsod right away at 2850mhz now. How high is too high with a venice, right now I'm at 1.6v, well... thats the highest any software is reading, I don't know where to check on the board with a dmm. CPUZ shows 1.55v bios shows 1.6v, so I'm saying 1.6v to be safe. I know you can go up to 1.7v safely, but I don't know how much my current cooling can take. I think I might beable to get 2800mhz by just burning in if it does indeed work, and maybe a slight bump in voltage.

[kryptobs2000]

Can anyone tell me where to measure the vcore at? I don't need pics, just tell me what to measure and the general area that it's at if you could.

[RebelsHaven]

Looking at the board mounted in a case, there are 3 identical inductors between the CPU and the ATX power connector. Vcore is the left hand lead, of any of the 3 inductors that is hard to reach because it's jammed up against the row of capacitors...

Or use this:

I measured Vcore and Vdimm using a Digital Multi-Meter. Here's what I got:

DFI nF3 Ultra-D
602 Bios
Venice 3200

Vcore

Default = 1.34V

CPU VID 1.50 = 1.44V
CPU VID 1.55 = 1.50V

Default +.1 = 1.38V
Default +.2 = 1.46V
Default +.3 = 1.51V
Default +.4 = 1.63V
Default +.5 = 1.67V
Default +.6 = 1.76V
Default +.7 - No Thanks!

1.55V +Default = 1.50
1.55V +.1 = 1.54V
1.55V +.2 = 1.64V
1.55V +.3 = 1.69V
1.55V +.4 = 1.83V
1.55V +.5 - No Thanks!
1.55V +.6 - No Thanks!
1.55V +.7 - No Thanks!

VDIMM

2.5V = 2.58V
2.6V = 2.69V
2.7V = 2.80V
2.8V = 2.91V
2.9V = 3.00V
3.0V = 3.10V
3.1V = 3.16V
3.2V = 3.17V

[Walrusbonzo]

umm,guys,i hope this program isnt accurate.note it saying core 2 vcore is 1.6!!??i mean,can both core temps be monitored like that?

cpu-z doesnt tho.it shows vcore 1.20 on both core..

thing is i am geting faily high temps at default 1.2 vcore(says bios) even?yes,im keeping it bottomed out till i figure my temps for sure.so,what ya think of accuracy of bios temps?thx..stock bios,big typhoon cooler..

oh,x2 4400+

erratic changes in bios temp too.it will sometimes dip to 16c?!

http://users.adelphia.net/~ace7/images/abcdef.jpg

http://users.adelphia.net/~ace7/images/abcdefg.jpg


I have exactly the same concerns. I'm currently running with only 1.28v VCore, even at stock speed my CPU temps idle show as 39C~. Yet I'm watercooling, the same cooler cooled my 3000+ Clawhammer down to 45C~ full load even though it was running at 2.5GHz with 1.85v!

Occasionally the CPU temp will just dip by 15C~20C. Having felt my radiator and the copper on my waterblock I think the lower temps it shows are probably the correct ones, but I can't be sure :(

[kryptobs2000]

Mine seems pretty accurate, with about 1.6v on my venice and an arctic freezer 64, it reads about 43c (Idle?) in the bios.

[Walrusbonzo]

Looking at the board mounted in a case, there are 3 identical inductors between the CPU and the ATX power connector. Vcore is the left hand lead, of any of the 3 inductors that is hard to reach because it's jammed up against the row of capacitors...

Or use this:

I measured Vcore and Vdimm using a Digital Multi-Meter. Here's what I got:

DFI nF3 Ultra-D
602 Bios
Venice 3200

Vcore

Default = 1.34V

CPU VID 1.50 = 1.44V
CPU VID 1.55 = 1.50V

Default +.1 = 1.38V
Default +.2 = 1.46V
Default +.3 = 1.51V
Default +.4 = 1.63V
Default +.5 = 1.67V
Default +.6 = 1.76V
Default +.7 - No Thanks!

1.55V +Default = 1.50
1.55V +.1 = 1.54V
1.55V +.2 = 1.64V
1.55V +.3 = 1.69V
1.55V +.4 = 1.83V
1.55V +.5 - No Thanks!
1.55V +.6 - No Thanks!
1.55V +.7 - No Thanks!

VDIMM

2.5V = 2.58V
2.6V = 2.69V
2.7V = 2.80V
2.8V = 2.91V
2.9V = 3.00V
3.0V = 3.10V
3.1V = 3.16V
3.2V = 3.17V


Just used a multimeter myself, the VDimm voltages are all bang on what you said :)

[kryptobs2000]

Oh, didn't notice that, not for me. I only checked at 2.9v which read 3.00v but at 3.2v I read 3.25v

[RebelsHaven]

If your 3.3V rail is high, the 3.1 and 3.2Vdimm readings will come out higher then the setting, just like the other Vdimm settings....

That's because the Vdimm is derived from the 3.3V rail on this board....

[Walrusbonzo]

OMG, am I doing something wrong??? I need to give my Venice 1.55v for it to even boot Windows at only 2.4GHz!!!!!! I'm using top watercooling gear too, I just don't believe my luck, have I got the crappest Venice 3000+ ever?

It's code is LBBLE 0520EPMW FYI. :(

[kryptobs2000]

OMG, am I doing something wrong??? I need to give my Venice 1.55v for it to even boot Windows at only 2.4GHz!!!!!! I'm using top watercooling gear too, I just don't believe my luck, have I got the crappest Venice 3000+ ever?

It's code is LBBLE 0520EPMW FYI. :(

You should probably read about the dfi NF3 before assuming it's your venice and posting in this thread. 1.55v = about 1.4-1.45v

[Walrusbonzo]

You should probably read about the dfi NF3 before assuming it's your venice and posting in this thread. 1.55v = about 1.4-1.45v

1.55v was using my Multimeter!

BIOS is set to 1.5v + 0.3v!

Sorry for not explaining before.

[Walrusbonzo]

So is anyone else getting higher than expected CPU temperatures?

[kryptobs2000]

oh, well that is wierd about your overclock then. Maybe it is just the worst venice there is lol. But I have heard of people getting high temp readings, I'm not one of them though, mine seem accurate enough. What bios are you using? I'm on the 629.

[Walrusbonzo]

629 also :)

I'm just learning to ignore the temps at the moment as I'm not at all convinced they're correct. If they are correct it must be a case of the core making extremely bad contact with the heatspreader. I may try removing the HS later as I did it for my 3000+ clawhammer too.

[kryptobs2000]

Yeah, ever since I moved from an nf7-s to an A64, I've gone through 4 boards, this is the first one that seems even somewhat right. I never check the temps anymore. I just know what is a safe voltage, if it gets too hot, it'll be unstable and I go by that.

edit: how much will removing the heatspreader likely gain on a venice you think? Is it worth it? 50mhz is enough for me to deam it worth it. But I don't wanna damage my chip, though I doubt it's too hard. People always said removing the shim on 9800/9700's was really dangerous and hard, but I did 3 no problem.

[terror_adagio]

629 caused problems for me, went back to 602.

The board defaults to DDR 333 for me. I cannot get DDR 400 at stock with 1T or 2T with 4x512. I'm suppose to be able to, but I can't.

But if I overclock the CPU to get to DDR 400, I can.

The temp readings are all wrong on this board, and it undervolts.

DFI needs to get on their ass and start fixing this and stop releasing BIOSs that are not fixing the real problems.

[Walrusbonzo]

What RAM slots is everyone using? I'm having a terrible time trying to get it Prime 95 stable. It keeps hard locking my PC.

Everything is setup to be error free in MemTest86+. After mucho testing, dropping CPU and RAM speed doesn't stop it from hard locking in Windows :(

Currently running with my RAM in the orange slots instead of the yellow ones, hoping this helps, and so far, it's stable. Fingers crossed.

[Walrusbonzo]

Back to the orange slots :/

[kryptobs2000]

Both slots are the exact same for me, I can't use 1T no matter what, but my only issue with this board is ram stability.

[Walrusbonzo]

I'm still unstable, but I think i'm almost at the end of the road :)

[terror_adagio]

Thought I would advise that there is apparently a "newer" official 705 BIOS out on DFI's site.

NF3ID705.zip (the previous BIOS was NF3UD705.zip) and some people are reporting better stable/performance with it.

[Nettwerk]

Good god this mobo is nutty. It decides to change bios settings when it wants, o/c's by ITSELF, cold boots, boots when it feels like it, decides it doesnt see my raid drives after a reboot EVEN WHEN I JUST BOOTED FROM IT 4 MINS AGO.. geezus :slapass: :stick:

[terror_adagio]

Are you using 705 BIOS.

No issues at all here.

[Nettwerk]

Yup 705 bios.. I swear I cleared cmos and went into the bios to just change my ram timings and set raid on and sata ports 3 and 4 on and reboot (everything else auto still) and go back to the bios and it sets my HTT to 269 from default 200?? So I laugh and change it back to 200 and save, exit, come back and it set it back to 269 HTT! So I change it again and it saves the third time.. lol

So I get everything to save finally and it saw my 2 harddrives on sata port 3 and 4, turn raid on those ports, tell it to boot from there first, then boom.. Doest see the drives on the next reboot... uhhgg! It was working fine yesterday and installed windows and everything on my 2 raided raptor drives... I cant figure this one out at all..

I went to change one of my ram timings in the bios (trrd or something) from 3 to 2 and it took 2 mins to finally post.. I got disgusted and just let it sit there cause I thought I had to clear cmos and enter in my ram timings that work again for the 2130319 time and jumped on my other computer and I heard it beep and post 2 mins later.. wtf did it just go take a piss break or something and decide it wanted to play now? :rolleyes:

[Giana]

^^ There is a 711 Bios actually released at the DFI Page ;) Perhaps i will try this later ;)

[Nettwerk]

^^ There is a 711 Bios actually released at the DFI Page ;) Perhaps i will try this later ;)


Yea just saw it im gonna flash as soon as I can find a working floppy disk lol

Edit: flashed and good news for me so far :) It actually sees my raid now! It actually saves my settings and posts after saving settings :banana: I caught it overclocking itself when I reset cmos after I flashed but changed it back to defualt and it saved.. I am now at 300x9 166 divider 2.5-3-3-6 stable again. They changed the tref value settings from reading (ex. 166us 3.9) to the number instead. so far so good!

[PoL]

Where can I measure vcore on this board?

Thanks! :)

[PoL]

Looking at the board mounted in a case, there are 3 identical inductors between the CPU and the ATX power connector. Vcore is the left hand lead, of any of the 3 inductors that is hard to reach because it's jammed up against the row of capacitors...


OK, sorry, I read this, but... Don't really know which one. Can somebody take a picture of it and where to put the voltimeter?

Thanks! :)

[PoL]

Doesn't matter, I found it.

Thanks anyway... :D :D

[msmolt]

Looking at the board mounted in a case, there are 3 identical inductors between the CPU and the ATX power connector. Vcore is the left hand lead, of any of the 3 inductors that is hard to reach because it's jammed up against the row of capacitors...

Or use this:

I measured Vcore and Vdimm using a Digital Multi-Meter. Here's what I got:

DFI nF3 Ultra-D
602 Bios
Venice 3200

Vcore

Default = 1.34V

CPU VID 1.50 = 1.44V
CPU VID 1.55 = 1.50V

Default +.1 = 1.38V
Default +.2 = 1.46V
Default +.3 = 1.51V
Default +.4 = 1.63V
Default +.5 = 1.67V
Default +.6 = 1.76V
Default +.7 - No Thanks!

1.55V +Default = 1.50
1.55V +.1 = 1.54V
1.55V +.2 = 1.64V
1.55V +.3 = 1.69V
1.55V +.4 = 1.83V
1.55V +.5 - No Thanks!
1.55V +.6 - No Thanks!
1.55V +.7 - No Thanks!

VDIMM

2.5V = 2.58V
2.6V = 2.69V
2.7V = 2.80V
2.8V = 2.91V
2.9V = 3.00V
3.0V = 3.10V
3.1V = 3.16V
3.2V = 3.17Vthere are many of us nf3 250 user on the street. thats been trying to find this info can you please show a pic of this location for vcore test points. thank you

[THunDA]

there are many of us nf3 250 user on the street. thats been trying to find this info can you please show a pic of this location for vcore test points. thank you

Hey msmolt.. ;)

The board he was testing this on was the NF3 Ultra-D.. (939 mobo)

DFI nF3 Ultra-D
602 Bios
Venice 3200



:toast:

[msmolt]

Hey msmolt.. ;)

The board he was testing this on was the NF3 Ultra-D.. (939 mobo)


:toast:Ahhhh Rats. ok sorry was hoping to find out this infermation on the nf3250 but guess still no luck sorry guys--- Michael---

[AsAsIn8eR]

I saw somewhere thoose measure points.I will edit my post and give you a link.

Sorry for the sig for the moment it is being resized

[msmolt]

I saw somewhere thoose measure points.I will edit my post and give you a link.

Sorry for the sig for the moment it is being resized Ha Ha look when i posted that lol what you mean a 100 post just to look at the for sale tread :slapass: ha ha heck all i needed was 2x512 of bh-5 or ch-5 or micron 5D-G lol maybe TccD if its a grate price

[i found nemo]

does it still have the conflict where in the games it freezes up??