[First Look] DFI 680i LT NF-680i LT SLI (56K Warning~~~)

Quote:

---

Originally Posted by virtualrain

So palese, what CPU were you clocking on the DFI and what settings did you try?

Also, what were you able to achive on the EVGA?

I'd like to know your experience on both boards in more detail (as painful as it may be for you to re-live! :p: ).

---

Xeon X3220, I tried all the GL settings Tony listed as well as the one he likes, All I want is a 3.1Ghz or 3.2Ghz OC with moderate FSB. Nothing works but what was funny was when I got to the end of my rope I pushed the clock speed to 3.6 out of frustration and for the first time I got to windows only to crash but at least it passed safe mode, so I slowly lowered everything but was not able to get pass safe mode again.

On the Evga Rev1 with the P27 bios I reached 3.2Ghz and FSB if I remember was something like 360. Also my cpu was set at 1.35v with no other volt adjustments.

Quote:

---

Originally Posted by Lood007

I'm sure Mr. Wu is still tweaking this as we speak. As long as he sticks to it, I believe we'll see some good Quad OC in the future.

---

Maybe they'll stick with it if more people are actually able to purchase one. The results seem decent enough for such an immature board. The layout and appearance IMO is awesome (far better then anything else I've seen). But none of it matters if no one can find one to purchase and when they do, have to spend $300 for it. Usually support & sales of a board go hand and hand so if the price drops and more people can get a hold of them then maybe they'll stay working on it.

So, here's an update on my overclocking adventures... (read back to post #879 for the background).

I started with these settings:

CPU VID: 1.45 (start at reasonable overvolt)
CPU Drive: Strong
DRAM Drive: Weak
DRAM Volts: 2.22 (just to keep RAM happy)
NB Core: 1.39 (start small)
SB Core: 1.62 (as per Tony's guide)
SB Dual: 1.52
LDT: 1.22 (no OC'ing here)
VTT: 1.40 (as per Tony/Raju's suggestion)
GTL0/1: 170
GTL2/3: 95
GTLNB: 85

Started OC'ing and running 30min. test with OCCT:

300FSB = OK
320FSB = OK
340FSB = OK
360FSB = OK
380FSB = Windows Doesn't Load

Increase CPU VID: 1.475; NB: 1.42
380FSB = Windows Doesn't Load
Roll back to 370FSB
370FSB = No Post - I assume there's a hole here back to 380FSB
380FSB = Windows Doesn't Load
Try tweaking GTL2/3 in steps of 2-3 points from 95 to 128
Try tweaking GTL0/1 up by 5
Increase CPU VID: 1.5; NB: 1.44
Try VTT=1.44 with above GTL values
Increase CPU VID: 1.55; NB:1.51
Try VTT=1.49 with above GTL values
Give up on 380... can't get Windows to load - must be a hole

Start 385FSB with CPU VID= 1.55, NB=1.51, VTT=1.44, GTL0/1=180, GTL2/3=100, GTLNB=90

385FSB = OK
395FSB = OK
400FSB = OK

Great! I'm basically back where I was last Friday (blog link). So I decide to run PRIME overnight... failed after 45min.

I just don't think OCCT is a good stability test... 30min is not long enough to assess true stability. I think it uses some form of PRIME code anyway based on the way I could see it cycle loading the cores just like different FFT sizes in PRIME.

So I've proven that my board/proc will boot and be somewhat stable with a variety of settings at 400FSB. It works at:

VTT=1.44; GTL0/1=180; GTL2/3=100; GTLNB=90
VTT=1.52; GTL0/1=180; GTL2/3=180; GTLNB=90

So I guess my processor is very flexible with GTL settings but nothing is rock solid stable.

I really don't know what to do to get a real stable (i.e. 4+hours IMHO) overclock with 380+ FSB. It just doesn't seem to be in the cards here.

LOL, u guys are really pushing me towads grabbing an EVGA when i get my quad... i had really high hopes for this board as all my previous DFI's have been great, but it seems like this one is alot of work for no extra gain.. at least as far as quads are concerned.

@ VR - You know you can set OCCT to run as long as you want and or infinitely, yes? If it's proof for others that you have a stable OC you may be spinning your wheels. Mayhaps stick with settings that pass 30 mins of OCCT and just start throwing everything you can at it to see if you encounter BSODS and lockups...If you can pass EVERYTHING but prime you may just want to take a break and just try using your rig for a couple of days...

when i get a chance today i'll try those settings with any necessary tweaks... if i can pull off a 400 x 8 i'd be fine with that for now...

everest has a system stability test that can stress all 4 cores do people ever use it for reliability testing?

Quote:

---

Originally Posted by OldGuy

@ VR - You know you can set OCCT to run as long as you want and or infinitely, yes? If it's proof for others that you have a stable OC you may be spinning your wheels. Mayhaps stick with settings that pass 30 mins of OCCT and just start throwing everything you can at it to see if you encounter BSODS and lockups...If you can pass EVERYTHING but prime you may just want to take a break and just try using your rig for a couple of days...

---

I guess so... perhaps I'm being overly demanding. It irks me that I can run prime endlessly at stock but it fails after 45min on OC. Is that a good OC? It use to be that 8+hrs PRIME stable was the measure of a good OC. The other thing with PRIME was that it was the ONLY test you needed to run... if you could pass it, you could be assured you wouldn't error out in anything else.

Quote:

---

Originally Posted by virtualrain

I guess so. It irks me that I can run prime endlessly at stock but it fails after 45min on OC. Is that a good OC? It use to be that 8+hrs PRIME stable was the measure of a good OC.

---

I'm from the 8+ hour school as well but at some point you have to question what is still valid moving forward and whether or not you want to keep hitting your head against the wall...take a breather ;)

Quote:

---

Originally Posted by OldGuy

I'm from the 8+ hour school as well but at some point you have to question what is still valid moving forward and whether or not you want to keep hitting your head against the wall...take a breather ;)

---

You have a point. Maybe after a new BIOS release it will be worth getting back at... :brick:

:toast:

C2D and 680i will be legacy hardware by the time this hits the UK.

P965 Infinity has been out for some time now and still it aint even in the UK. Not, it's been in stock and sold out. It ain't never been in stock ONCE.

Pathetic distribution.

Quote:

---

Originally Posted by Mekrel

C2D and 680i will be legacy hardware by the time this hits the UK.

P965 Infinity has been out for some time now and still it aint even in the UK. Not, it's been in stock and sold out. It ain't never been in stock ONCE.

Pathetic distribution.

---

Be carefull what you wish for :D

well it sold out at newegg... but took much longer than i expected. hopefully oskar will be devoting lots of time to the BIOS.

Quote:

---

Originally Posted by palese

Be carefull what you wish for :D

---

Hey Palese... I still want to hear your story (more specifics) with the DFI and what you get on the EVGA.

Quote:

---

Originally Posted by virtualrain

Hey Palese... I still want to hear your story (more specifics) with the DFI and what you get on the EVGA.

---

Well for the DFI I used Tony's guide as well as your settings, I can not get past safe mode, there must be something in bios that I am overlooking that needs to be disabled :confused: This is the first board "and I do have a few" that I can not get into windows with on an OC.

If I could just get into windows then the tweaks would not be that big a deal, but safe mode is all I get. Stock is not an issue at all so I know all my hardware is compatible.

Have you tried the new Evga rev1 ? If not you owe it to yourself to do so, there is not a lot to say about it other then it is by far the easiest board out there to OC.

Extreme OCing on the Evga would require robust cooling , but moderate OC speeds of say an E6600 or X6600 @ 3.0Ghz to 3.4Ghz air is fine with very minor Volt adjustments.

Quote:

---

Originally Posted by virtualrain

380FSB = Windows Doesn't Load

Increase CPU VID: 1.475; NB: 1.42

---

Well... i guess this is proof, as if we needed it, that not all chips will react the same. using your base settings i starting clocking up. was a little nervous around 380 from your experience but i booted windows fine @380 and ran everest stress-test 10 mins to make sure it wasn't a fluke.

continued on to 390 and finally 400 with absolutely no changes to the settings whatsoever. my CPU is even under-volted at idle through droop by .04v

i'll do some stability testing now... i'm sure you're interested in Prime results. used that a lot back in the day with my abit board so i'll give it a go!

i took a screenshot for posterity...

http://i168.photobucket.com/albums/u...001/400fsb.jpg

Quote:

---

Originally Posted by Jobeo

Well... i guess this is proof, as if we needed it, that not all chips will react the same. using your base settings i starting clocking up. was a little nervous around 380 from your experience but i booted windows fine @380 and ran everest stress-test 10 mins to make sure it wasn't a fluke.

continued on to 390 and finally 400 with absolutely no changes to the settings whatsoever. my CPU is even under-volted at idle through droop by .04v

i'll do some stability testing now... i'm sure you're interested in Prime results. used that a lot back in the day with my abit board so i'll give it a go!

i took a screenshot for posterity...

http://i168.photobucket.com/albums/u...001/400fsb.jpg

---

More importantly can you show your bios settings, I cant get this POS to boot to windows with any OC:fact:

Sweet... but let's see 400x9! :D I could hit 3200MHz without changing any settings either... but getting to 3.6GHz, might be another story. :p:

At any rate, you've given me a new idea... I'm going to try 400-450x8 to see if I can get a more stable OC at 3.5 to 3.6GHz on the x8 multi.

Quote:

---

Originally Posted by palese

More importantly can you show your bios settings, I cant get this POS to boot to windows with any OC:fact:

---

lol... i feel your pain. just gotta stick with it. the settings i'm using are right from VR's post. except for the CPU vid.

although i can boot windows and everything runs fine its neither prime nor OCCT stable right now. fails both. as far as 3.6ghz with an x9 goes we'll see about that ! ! lol. i don't have the cooling for it right now. at 3.2ghz i'm hitting around 70C peak.

but i have some parts coming tomorrow which should help with that. and if heat really becomes and issue i can break out meh vapochill but i don't enjoy working with it.

http://i168.photobucket.com/albums/u...X001/bios2.jpg http://i168.photobucket.com/albums/u...X001/bios1.jpg

My God im in Windows @ 3.1 :banana: :woot:

Jobeo that combo worked for me thanks, and a big thanks to virtualrain as well.

Quote:

---

Originally Posted by palese

My God im in Windows @ 3.1 :banana: :woot:

---

:eek:

:)

Quote:

---

Originally Posted by palese

My God im in Windows @ 3.1 :banana: :woot:

Jobeo that combo worked for me thanks, and a big thanks to virtualrain as well.

---

I hope you mean you are in Windows at 3.1GHz and not Version 3.1 :eek: :D

Glad to hear you making progress!

For fun, I maxed voltages for a suicide run and was able to boot once into Vista x64 (a stability test on it's own) at 3.8GHz (425x9)! Not stable at all and could only get in at one set of GTL settings and even then only 1 in 3 times and doing anything would hang it. I was able to get CPU-Z and Speedfan going long enough to grab a screen :D

http://members.shaw.ca/virtualrain/3800mhz.png

CPU VID = 1.6
NB V = 1.65
SB V = 1.62
VTT = 1.6
GTL1 = 200
GTL2 = 170
GTLNB = 100

:eek: Don't try this at home! :eek:

@ Jobeo, I notice you are running unlinked/auto at 1600/800... doesn't linked/sync'd give you the same result (1600/800)? Any reason why you went with that?

Also, I see you OC'd the PCIe bus out of the gate... does that help, hinder, neither or unknown with respect to overall OC stability?

Quote:

---

Originally Posted by virtualrain

@ Jobeo

---

i went unlinked cause i had more success with it last week. ultimately i'd like to bump that to 1250 which is what my memory is certified capable of ...

as far as the PCIe clocks go i should probably just bump that down. i read a report somewhere that it barely even affects FPS anyways... i'm not sure what effect it has on an OC but its probably a good idea to bump it back so there's no chance for it to interfere...

Think i can live with it...for now

http://img517.imageshack.us/img517/2...0315hh0.th.jpg http://img452.imageshack.us/img452/2...3152ce0.th.jpg

Not bad, hopefully we'll see some BIOS improvements within the coming weeks.

Quote:

---

Originally Posted by virtualrain

I hope you mean you are in Windows at 3.1GHz and not Version 3.1 :eek: :D

Glad to hear you making progress!

For fun, I maxed voltages for a suicide run and was able to boot once into Vista x64 (a stability test on it's own) at 3.8GHz (425x9)! Not stable at all and could only get in at one set of GTL settings and even then only 1 in 3 times and doing anything would hang it. I was able to get CPU-Z and Speedfan going long enough to grab a screen :D

http://members.shaw.ca/virtualrain/3800mhz.png

CPU VID = 1.6
NB V = 1.65
SB V = 1.62
VTT = 1.6
GTL1 = 200
GTL2 = 170
GTLNB = 100

:eek: Don't try this at home! :eek:

---

:eek: :eek: :eek: That was not on air right:D

Quote:

---

Originally Posted by Jobeo

i went unlinked cause i had more success with it last week. ultimately i'd like to bump that to 1250 which is what my memory is certified capable of ...

as far as the PCIe clocks go i should probably just bump that down. i read a report somewhere that it barely even affects FPS anyways... i'm not sure what effect it has on an OC but its probably a good idea to bump it back so there's no chance for it to interfere...

---

On Asus boards the PCIE frequency bumps were needed for OC to keep the FX cards stable while ocing.

Another review up.

http://www.hothardware.com/articles/...80i_LT/?page=1

Quote:

---

Originally Posted by palese

:eek: :eek: :eek: That was not on air right:D

---

No... CPU cooled by water (Apogee GTX / Thermochill 120.2 / DDC Ultra)

Quote:

---

Originally Posted by palese

Another review up.

http://www.hothardware.com/articles/...80i_LT/?page=1

---

I notice their comments on OC'ing echo alot of our own...

Quote:

---

Our sample board was able to hit 1800 MHz FSB at stable levels, a solid overclock for a 1066 MHz FSB based dual-core processor, although not quite to the stratospheric levels which were claimed to be possible. DFI certainly does provide the tools for heavy overclocking, and their custom cooling systems do an excellent job of keeping the chipset cool. However, every board will overclock to different levels, ours simply didn't fare as well as expected. For the record, the Nvidia reference / eVGA nForce 680i LT reference board we tested a few weeks back was able to overclock slightly further, hitting up to 1933 MHz, although it too was only stable in the 1800-1866 MHz FSB range. However, we did find overclocking to be far easier on the eVGA platform compared to our DFI board.

---

I'm happy to report that my quest for an 8hr PRIME stable overclock at 3.6GHz on my X3220 has been fulfilled!!! Last night I ran some new GTL settings with very aggressive voltages and was rewarded! Finally!

:banana:

I'll post BIOS screens with settings after I get a bit of work done :)

For now, here's proof...

http://members.shaw.ca/virtualrain/primestable36.png

Quote:

---

Originally Posted by virtualrain

I'm happy to report that my quest for an 8hr PRIME stable overclock at 3.6GHz on my X3220 has been fulfilled!!! Last night I ran some new GTL settings with very aggressive voltages and was rewarded! Finally!

:banana:

I'll post BIOS screens with settings after I get a bit of work done :)

For now, here's proof...

http://members.shaw.ca/virtualrain/primestable36.png

---

Congrats :woot: :toast:

Nicely done! At least you have been rewarded for all your hard work in the end. What do u recon stabilised it? Just chucking lots of voltage at the chip, or was it to do with the GTL settings afterall?

Question on the northbridge heatsink: ARe you guys taking the white square sticker/pad off to get better contact with the heatsink or just leaving them both on?

Quote:

---

Originally Posted by vincey2kr1

Question on the northbridge heatsink: ARe you guys taking the white square sticker/pad off to get better contact with the heatsink or just leaving them both on?

---

I learned from my Tuniq Tower never use a stick pad , I change boards alot and when you try to remove the pad its a mess.

Quote:

---

Originally Posted by MikeMK

Nicely done! At least you have been rewarded for all your hard work in the end. What do u recon stabilised it? Just chucking lots of voltage at the chip, or was it to do with the GTL settings afterall?

---

I think it was definitely both Vcore and GTL that was holding me back... I couldn't have reached this stable OC without tweaking both.

Originally, I stayed away from high Vcore because it was making the board shutdown. I didn't realize until yesterday that this was because I was exceeding the thermal shutdown setting in the DFI BIOS (nice feature!). I thought it was just crashing. As you all know, a Quad on high OC running small FFT's generates a ton of heat - even on good water. To compound this problem of the board shutting down, the BIOS thermal sensor is always 10-15-deg above CoreTemp so that's why I was tripping the thermal shutdown. Once I realized this and reset the shutdown threshold, I could use higher Vcore.

Another issue is that you really need to overvolt to get good voltages as what you set is NOT what you get. Set your voltages, reboot back to BIOS and look at the values at the bottom of the Voltages setting screen. As you get to more extreme voltages, their reluctance to track your settings seems to increase.

My Board definitely uses some strange GTL settings to be absolutely stable but I could boot at 3.6GHz a very large range of GTL settings. Getting it stable takes some tweaking though. When I was playing around trying to boot Vista x64 at 3.8GHz last night, I tried every GTL setting I could and only a couple would allow Vista to fully load. I could actually get further into the startup routine before it crashed as the GTL values were tweaked. It was a great way to see what settings were warmer/colder so to speak.

Final settings:

http://members.shaw.ca/virtualrain/primestableoc001.jpg

There's definitely something not right about the droop on the NB core voltage.

There's a bit more info on my blog if you are interested and I've also updated the "Tips, Tricks, and Known Issues" article... http://virtualrain.blogspot.com/

anyone know if the danger den maze4 chipset block would work on this?

Quote:

---

Originally Posted by whoodiestyle

anyone know if the danger den maze4 chipset block would work on this?

---

They made one for the DFI RD600 board that should fit this board... http://www.dangerden.com/store/produ...&cat=46&page=1

Quote:

---

Originally Posted by virtualrain

I'm happy to report that my quest for an 8hr PRIME stable overclock at 3.6GHz on my X3220 has been fulfilled!!! Last night I ran some new GTL settings with very aggressive voltages and was rewarded! Finally!

:banana:

I'll post BIOS screens with settings after I get a bit of work done :)

For now, here's proof...

http://members.shaw.ca/virtualrain/primestable36.png

---

Awesome. What batch?

Quote:

---

Originally Posted by virtualrain

They made one for the DFI RD600 board that should fit this board... http://www.dangerden.com/store/produ...&cat=46&page=1

---

virtualrain, question for you I down loaded OCCT and it shows my X3220 as a clovertown but yours shows as a Kentsfield.:confused:

Quote:

---

Originally Posted by nugzo

Awesome. What batch?

---

L640F058

@palese... Mine shows as Clovertown on OCCT as well.

Looks good... but that is indeed one hell of alot of droop on the NB. Do u recon it could be reporting incorrectly? Dont think ive ever seen so much droop.

BTW, has anybody heard anything about the reported in compatibility with Enermax Galaxy PSU's? I really need to know if this board works ok with them before i jump on one.

Quote:

---

Originally Posted by MikeMK

Looks good... but that is indeed one hell of alot of droop on the NB. Do u recon it could be reporting incorrectly? Dont think ive ever seen so much droop.

BTW, has anybody heard anything about the reported in compatibility with Enermax Galaxy PSU's? I really need to know if this board works ok with them before i jump on one.

---

It seems to be case by case as some run fine of the same model and some do not.

Bit of a gamble then. Might just have to suck it and see... switch to an EVGA if it doesnt work.

Quote:

---

Originally Posted by MikeMK

Bit of a gamble then. Might just have to suck it and see... switch to an EVGA if it doesnt work.

---

I own both boards now and I could not recomend the DFI over the Evga Rev1 at this point, at least not with a quad.

hmmm... yes, well thats what it will be for. I plan on grabbing a QX6700 to mess around with shortly, and from the results ive seen, my IN9 isnt really up to the job of getting the best out of quad core.

It is a tough decision. Past DFI's have served me well, and I do wonder if future bioses will bring this board up to another level.

Quote:

---

Originally Posted by palese

I own both boards now and I could not recomend the DFI over the Evga Rev1 at this point, at least not with a quad.

---

I would really like to try an EVGA board now to see if I could do the same OC or better... In my case, with fairly radical GTL settings required to get stable, I wonder if I would be worse off with a board like EVGA which has one-size-fits-all GTL?

I guess I could buy one from a local retailer, try it out and pay the restocking fee to return it within the alotted time.

Anyone have an EVGA A1/T1 they want to lend me?! :D :p:

Quote:

---

Originally Posted by MikeMK

hmmm... yes, well thats what it will be for. I plan on grabbing a QX6700 to mess around with shortly, and from the results ive seen, my IN9 isnt really up to the job of getting the best out of quad core.

It is a tough decision. Past DFI's have served me well, and I do wonder if future bioses will bring this board up to another level.

---

Oh im sure this DFI will see its day but even after you do boot into windows and find something that's stable, its a very lackluster experience. Very slow boot ups as well, I guess if I was not running my quads on the Evga board before installing this DFI I would have been fine.

Anyone able to run a Thermaltake toughpower1000w with this board?

So is this a good board for qx6700 _+}?

Has anyone tested 4x1GBs of Corsair 6400C3DFs with this board? Those modules aren't in memory QVL on the DFI website.

I apologise if someone has already posted about this. I didn't have time to read through 30 pages of info at work. Thanks in advance..

I pulled out my Fluke DMM tonight and did a bit of measurement on some key voltages...

CPU VID:

BIOS...
Set: 1.6000V
BIOS Reading: 1.53V
Fluke DMM: 1.54V

Windows (Idle)...
Speedfan/SmartGuardian: 1.55V
Fluke DMM: 1.56V

Windows (PRIME Small FFT)...
Speedfan/SmartGuardian: 1.49V
Fluke DMM: 1.50V

Overall, BIOS readings and Speedfan/SmartGaurdian are close enough to actual Voltages to consider them legit.

VDimm:

Set: 2.34V
BIOS: 2.27V
SmartGaurdian: 2.27
DMM: 2.28V

Again, BIOS readings are close to actual at least when not under load.

NB:

Since I don't have a DMM measurement point for the NB, I just checked a variety of settings and BIOS/SmartGuardian readings...

Set / Get

1.60 / 1.47
1.57 / 1.45
1.54 / 1.42
1.51 / 1.39
1.48 / 1.37

There is a consistent droop of 0.12V on the NB... If the BIOS reading is correct (as it was with Vcore and Vdimm above) then this is very poor... keep this in mind when overclocking to a high FSB.

This question has probably been answered before but I'm not going back and reading the entire thread.

How does the DFI 680i LT stack up against Evga's?

Quote:

---

Originally Posted by Yay for Cheese

This question has probably been answered before but I'm not going back and reading the entire thread.

How does the DFI 680i LT stack up against Evga's?

---

I think the EVGA is alot more mature board. Been out for a while and with the quadcore GTL revision (A1 and T1) boards are alot easier to overclock.... Cant get my Q6600 to boot at 1333fsb thinkin about going back to my ASus p532-e sli .:shakes:

Quote:

---

Originally Posted by virtualrain

I pulled out my Fluke DMM tonight and did a bit of measurement on some key voltages...

CPU VID:

BIOS...
Set: 1.6000V
BIOS Reading: 1.53V
Fluke DMM: 1.54V

Windows (Idle)...
Speedfan: 1.55V
Fluke DMM: 1.56V

Windows (PRIME Small FFT)...
Speedfan: 1.49V
Fluke DMM: 1.50V

Overall, BIOS readings and Speedfan are close enough to actual Voltages to consider them legit.

VDimm:

Set: 2.34V
BIOS: 2.27V
DMM: 2.28V

Again, BIOS readings are close to actual at least when not under load.

NB:

Since I don't have a DMM measurement point for the NB, I just checked a variety of settings and BIOS readings...

Set / Get

1.60 / 1.47
1.57 / 1.45
1.54 / 1.42
1.51 / 1.39
1.48 / 1.37

There is a consistent droop of 0.12V on the NB... If the BIOS reading is correct (as it was with Vcore and Vdimm above) then this is very poor... keep this in mind when overclocking to a high FSB.

---

Great work virtualrain, thanks for the time and effort you have put into this board to make our life with it a hell of a lot less frustrating. :up:

Quote:

---

Originally Posted by worthIT

Has anyone tested 4x1GBs of Corsair 6400C3DFs with this board? Those modules aren't in memory QVL on the DFI website.

I apologise if someone has already posted about this. I didn't have time to read through 30 pages of info at work. Thanks in advance..

---

I can test it with 2GB tonight after work for you, I lost 3 sets on the first Evga board before the revision:shakes: I have been running cellshocks since then and never went back to running the corsairs again, but they are new in the box and I am now curious myself how they will fare on the DFI.

Just got my DFI in, Installed it, and guess what?

DOA. Will not post.


I am not a happy camper right now.

Quote:

---

Originally Posted by Spazilton

Just got my DFI in, Installed it, and guess what?

DOA. Will not post.


I am not a happy camper right now.

---

you wanna give any more details ?? what bios post code ??

Been getting BSOD's in LIneage II pulling my hair trying:mad: to figure out what it could possible be... Finally I swapped everything back (q6600,4gigs Patriotextreme, Thermaltake toughpower 1000w, 8800GTS 320mb) to my trusty Asus p532-e SLI and a seagate 320gig perp drive and no more stupid npkcrypt.sys errors:clap:

Waiting this long for this DFI board hoping all the bugs/kinks would be worked out of it is just a big Dissapointment :shakes: Especially for the price >< Hoping i can get a refund I need something that works Now *sigh*.

Couldnt even get a 1333 FSB overclock on this DFI 680 with myQ6600, First try 1333fsb Overclock on the ASus p532-e with No problems

FYI, I put a post up in the memory/bandwidth thread asking for input on my memory performance on this board if you guys are interested and/or can provide some advice...

http://www.xtremesystems.org/forums/...d.php?t=144822

Quote:

---

Originally Posted by palese

I can test it with 2GB tonight after work for you, I lost 3 sets on the first Evga board before the revision:shakes: I have been running cellshocks since then and never went back to running the corsairs again, but they are new in the box and I am now curious myself how they will fare on the DFI.

---

That would be much appreciated. I've also killed two kits of this RAM on my STRIKER. 2GBs I can get running no problems, but 4GBs never sat well with this board and I have to loosten off pretty much everything just to get the board stable.

If 4GBs with this board is an option, I might try that, otherwise I will just change my board to the EVGA just to end the pain..

Don't give up this board so soon. Took me a long time to get things stable and to get a feel for what the board likes and doesn't like.

I have gotten my ram up to 950MHz 1T, 4-4-4-8, 2-1-1-7-7.8us on this board. Mem bandwidth is 5600MB in memtest86+!

As obvious as this should be, it probably still needs to be said in plain words. If you do not enjoy tweaking, you should not be looking at this board. If you want plug and play, instant overclock, look elsewhere. The multitude of settings here require a time investment, and no small one at that, to master and get a feel for as Lood007 said. But, if you are willing to invest that time, this board will reward you.

Quote:

---

Originally Posted by century child

As obvious as this should be, it probably still needs to be said in plain words. If you do not enjoy tweaking, you should not be looking at this board. If you want plug and play, instant overclock, look elsewhere. The multitude of settings here require a time investment, and no small one at that, to master and get a feel for as Lood007 said. But, if you are willing to invest that time, this board will reward you.

---

Sorry but that spin does not wash anymore, the issues with this board is not lack of user know how, it starts with poor instruction from DFI
Example: GTL But aside from that this 680i offering from DFI offers no better results in the end then any other 680i offering out there, so tweak your brains out all you want.

Make it a point to read the review on Hot Hardware as they could not reach the FSB numbers that DFI claims to have reached, Yes you could just say the Hot Hardware folks did not understand the time and tweaking you need to put into this board to reach Mecca but Hot Hardware enlisted the help of DFI engineers to reach the goal but DFI could not make it happen as well. :ROTF:

I have no doubt that in time with updated bios releases that this 680i could stand out above the rest of the pack but it maybe to little to late and what is wrong with plug and play easy OC anyways:shrug: Who cares how you get there;)

Quote:

---

Originally Posted by palese

Make it a point to read the review on Hot Hardware as they could not reach the FSB numbers that DFI claims to have reached, Yes you could just say the Hot Hardware folks did not understand the time and tweaking you need to put into this board to reach Mecca but Hot Hardware enlisted the help of DFI engineers to reach the goal but DFI could not make it happen as well. :ROTF:

---

But nowhere did DFI state that EVERY board would be able to reach 2000 MHz FSB. They claimed it was possible, and showed as such in their demonstrations.

I am not here to defend them blindly, as I do own and like many other motherboards, but I like the tweakability of this board... and I take marketing with a grain of salt... every board, CPU, RAM set, video card, etc is different, and your results will never match exactly what someone else has gotten.

regardless of the tweak factor... the PWM overheating issue cannot be overlooked. DFI designed a board for overclockers, a LANPARTY board, and doesn't expect people to be verifying their system prime stable ??

its a serious design flaw and hopefully one they address. the PWM heatsink is clipped on the board so they could release a better solution for us early adopters.

Extra tweacking on the Nforce series of DFI was justified by the extra performance that you could get compared to the other boards.Thats why users loved them. They had a long and hard way with them but in the end of the day they got rewarded for that.

Nowadays in the conroe era there are thousands of boards that can do same or better performance with less effort. After all what counts is the money and time you spend to reach your aim. I agree that the procedure has all the fun to OC but it should be interdependent to the results you get, otherwise it would seem stupid spending so much hour and effort when somebody else can make it easier and faster than you.

I love DFI as I own two boards of them. But then again I see other manufactures with less experience than DFI making it better nowadays.:owned:

Well, all I have to say so far, Ive had this board for just over 24hrs and absolutly love playing with it :up:

I'm not saying that other 680i boards are better or worse and this one is probably avg in terms of performance, I just haven't got that far with it.

I do have som odd inconsistancies regaurding different programs and how they report voltage and temps (I'll let the pic do the talking) What should I believe?

This board has achieved me a more stable overclock so far than my P965-s 3200MHz stable, 3400MHz at these settings would fail orthos after 3 min.

I will be pushing it harder as long as it passes 8hrs orthos.. got 4.5hrs so far

http://a731.ac-images.myspacecdn.com...e8680d1b7a.jpg

Quote:

---

Originally Posted by xTr33m sammich!

What should I believe?

---

i trust the smart guardian numbers... they are very similar to the levels reported in the BIOS which are the most accurate thing we have to go by.

Quote:

---

Originally Posted by palese

Sorry but that spin does not wash anymore, the issues with this board is not lack of user know how, it starts with poor instruction from DFI
Example: GTL But aside from that this 680i offering from DFI offers no better results in the end then any other 680i offering out there, so tweak your brains out all you want.

Make it a point to read the review on Hot Hardware as they could not reach the FSB numbers that DFI claims to have reached, Yes you could just say the Hot Hardware folks did not understand the time and tweaking you need to put into this board to reach Mecca but Hot Hardware enlisted the help of DFI engineers to reach the goal but DFI could not make it happen as well. :ROTF:

I have no doubt that in time with updated bios releases that this 680i could stand out above the rest of the pack but it maybe to little to late and what is wrong with plug and play easy OC anyways:shrug: Who cares how you get there;)

---

Blow it off as spin if you must but, based on my personal experiences with this board and those of other forum users, it's the truth whether you want to hear and/or believe it or not. You obviously belong to the group that doesn't so much enjoy endless hours of tinkering with BIOS settings and there is nothing wrong with that. Nor was I trying to imply there is something wrong with plug and play easy OC. I was simply making the statement that this board, as with most other boards in the history of the LanParty series, are best suited for people who enjoy and have time for endless tweaking and experimenting.

Quote:

---

Originally Posted by century child

Blow it off as spin if you must but, based on my personal experiences with this board and those of other forum users, it's the truth whether you want to hear and/or believe it or not. You obviously belong to the group that doesn't so much enjoy endless hours of tinkering with BIOS settings and there is nothing wrong with that. Nor was I trying to imply there is something wrong with plug and play easy OC. I was simply making the statement that this board, as with most other boards in the history of the LanParty series, are best suited for people who enjoy and have time for endless tweaking and experimenting.

---

I'm going to have to agree with you here.

As it stands right now, this DFI board is already beating my BX2 in terms of Quad FSB overclocking. It is leaps and bounds better at Quads than my IN9 32X.

With some minor adjustments to LDT, and the GTL settings, I have already hit 3.2 Prime Stable and climbing on STOCK Vcore. BX2 required 1.4125 with Enchanced Power Slope, IN9 refused to boot periord at this speed regardless of voltage(Board topped out at 300 FSB on quads regardless of settings.

This board has massive potential.

Quote:

---

Originally Posted by Jobeo

i trust the smart guardian numbers... they are very similar to the levels reported in the BIOS which are the most accurate thing we have to go by.

---

Yes... Jobeo is correct. Read back to Post #952.

As far as DFI vs. EVGA or others, I think if this boad had come out at the same time as the EVGA board, there would probably be very few EVGA users right now... that board had so many problems initially. (Although I suspect that DFI probably avoided teething issues with this chipset by delaying so it's not black and white).

However, the fact is that DFI is very late to the party and EVGA has done an admirable job of making an outstanding overclocking board with it's revs and bios updates (from what I've read). DFI is playing catch-up in that regard.

The EVGA board (these days) has the benefit of a large installed base and a huge community which makes it easier for noobs as well. With the DFI board being so new, that installed base and collective wisdom hasn't yet developed, but will over time.

At any rate, given that my stable quad overclock requires fairly radical GTL settings, I suspect I would not get the same level of overclock on a board that did not provide GTL tuning. I think those guys that get the same overclock on both boards are just fortunate to have a "typical" CPU that fits the profile to which the EVGA board was designed. If you don't have such a CPU, then what? You either get a board like the DFI or you suffer with a low overclock. Remember, Not all EVGA quad owners are hitting 3.6GHz... who knows what they might achieve if they put their quad in a DFI and spent time tuning it?

One final thought... sadly, by the time DFI gets this board rocking and the collective user community has it nailed, it will probably be obsolete.

question
How hot is too hot for the NB on this board? Virtualrain, it seems youre NB is running somewhere between 40 and 50c, what are you cooling your NB with? Past 440fsb I need to crank up NBv to 1.6v or better, results in my NB running at approx 80c:shocked: I don't like that..

I am going to try to re-arrange my rig a little, the ultra120 extreme with 2x 120mm and 2x8600gt vid cards it gets a little cramped and can't get a fan on the NB... holding a fan pointed in the general direction of the nb drops it to approx 75c.. still too hot for me, but better.

Quote:

---

Originally Posted by xTr33m sammich!

question
How hot is too hot for the NB on this board? Virtualrain, it seems youre NB is running somewhere between 40 and 50c, what are you cooling your NB with? Past 440fsb I need to crank up NBv to 1.6v or better, results in my NB running at approx 80c:shocked: I don't like that..

I am going to try to re-arrange my rig a little, the ultra120 extreme with 2x 120mm and 2x8600gt vid cards it gets a little cramped and can't get a fan on the NB... holding a fan pointed in the general direction of the nb drops it to approx 75c.. still too hot for me, but better.

---

Wow... that seems way high... Perhaps try remounting the NB sink.

I just have a basic 120mm fan leaning up against the NB cooler. I guess it's pretty effective since it cools the top and base together. I'm working with it open on the bench though so that helps as well. I'll be water cooling it when it eventually goes in the case.

Here's pics... http://virtualrain.blogspot.com/2007...-dfi-680i.html

Quote:

---

Originally Posted by xTr33m sammich!

results in my NB running at approx 80c:shocked: I don't like that..

---

holy croaks ! !? ! !! ! ! i would shut that thing off quick if i was hitting 80c. i can't believe it's even working at those temps. raja recommends the NB temps not go high than something like 48-50c and you are well past that.. wow.

do you mind posting pics of your MB with the ultra extreme mounted ? i'd like to see what kinda space you're dealing with. i have an ultra in the mail so i'll be going down that same road soon.

Don't laugh

sryy about the crude drawing, this is the best I can think of right now, with both fans pushing air into the HS, it creates good airflow through the NB HS, 5c higher cpu core temp vs one pushing and the other pulling, under load at 3.5GHz, No good.
I re-seated the NB HS and temps have dropped approx 30c down to 48-53c, much better :P
I will post some pics later, have to go get the kids...

BTW, I was running orthos for 1.5hrs with FSB@440, NB@80c, no issues, scary? yes...

http://a11.ac-images.myspacecdn.com/...3c806038da.jpg

Quote:

---

Originally Posted by xTr33m sammich!

Don't laugh

sryy about the crude drawing, this is the best I can think of right now, with both fans pushing air into the HS, it creates good airflow through the NB HS, 5c higher cpu core temp vs one pushing and the other pulling, under load at 3.5GHz, No good.
I re-seated the NB HS and temps have dropped approx 30c down to 48-53c, much better :P
I will post some pics later, have to go get the kids...

BTW, I was running orthos for 1.5hrs with FSB@440, NB@80c, no issues, scary? yes...

http://a11.ac-images.myspacecdn.com/...3c806038da.jpg

---

Sounds like a bad mount to me. Mine doesnt go over 45c or so

I have run into a new issue however.


It seems CPU throttling is somehow tied to VRM temps. Even with the CPU under 65c in core temp, the CPU starts to throttle its self. This is on water.

Once I place active cooling on the VRM it stops, take it away and it starts right back up.

Still trying to figure out what would be an effective VRM cooling solution over stock.

Quote:

---

Originally Posted by Spazilton

Still trying to figure out what would be an effective VRM cooling solution over stock.

---

lol that makes two of us... let me know if ya come up with anything

i don't understand your drawing... where do you have fans and where are they blowing ??

Quote:

---

Originally Posted by virtualrain

Yes... Jobeo is correct. Read back to Post #952.

---

I would trust the BIOS program for voltages, but temperatures I would trust CoreTemp.

Speedfan comes up with some ridiculous voltages sometimes and the VID on CoreTemp is the 'recommended' voltage for your processor. Note: it doesn't change no matter what you do to the settings in the BIOS.

The cpu temperature out of your BIOS program isn't read from the same sensor as your CoreTemp program so technically, both are right.

Quote:

---

Originally Posted by worthIT

The cpu temperature out of your BIOS program isn't read from the same sensor as your CoreTemp program so technically, both are right.

---

what sensor does the bios program read from ? i'd like to trust coretemp for my therms cause its reporting about 6c lower than smart guardian...

intel rates my proc as max 65c which i can hit with smart guardian where coretemp reports none of my cores higher than 58c... which should i trust ?

Quote:

---

Originally Posted by Jobeo

intel rates my proc as max 65c which i can hit with smart guardian where coretemp reports none of my cores higher than 58c... which should i trust ?

---

The Bios sensor reads from a sensor based in your CPU socket whereas CoreTemp reads from the temperature sensor inside each core on your CPU. As far as I know, CoreTemp is the only program that can read from the temperature sensors inside your physical cores on your CPU, and is therefore classed as more accurate.

Everest also reports the core temp and it is spot on with coretemp...

Quote:

---

Originally Posted by worthIT

The Bios sensor reads from a sensor based in your CPU socket whereas CoreTemp reads from the temperature sensor inside each core on your CPU. As far as I know, CoreTemp is the only program that can read from the temperature sensors inside your physical cores on your CPU, and is therefore classed as more accurate.

---

Here's what I've figured out about monitoring on this board...

BIOS, Speedfan, and SmartGaurdian all pull monitoring values from the ITE SuperIO Chip the IT8712F which includes voltages and a few temps (CPU, NB, and SB).

Note for those who don't already know... the temp in BIOS called "PWM Area" is mislabled... this is actually the NB temp.

Like many others have said, the CPU temp pulled from the IO chip should not be trusted... CoreTemp rules. Both Everest and Speedfan also pull the same data from the same registers as CoreTemp so you can use Speedfan or Everest to do the same thing. With Speedfan, you simply need to add a 15-deg C offset to the reported values to make them match with Coretemp (it's built-in offset is not correct but the values from the registers are).

Speedfan can see Vcore, the HT Voltage, and the NB Voltage but initially displays them as Vcore1, Vcore2, and 3.3V... simply rename them with the proper labels.

Unfortunatly Speedfan current doesn't seem to report the VDimm or SB Voltage, but not a big deal. It also can't control the fans which is a shame... you need SmartGaurdian for that (why is it so damn ugly!).

Hopefully a future version of Speedfan will offer more comprehensive monitoring of this board. I'm logging an issue with the author now.

That's about it for monitoring on this board that I know of.

For water cooling the PWM area, there is no current block that will work with the mounting hole spacing provided (almost all are either way too long or too short).

So I'm thinking of buying one of these and adding a couple of mounting ears to it so I can fasten it down. It's a bit longer than the stock heat sink (65mm vs 50mm for stock) but I think it would work nicely with the mounting hole most inboard and then I could use the ATX motherboard screw down point near the edge of the board to fasten the other end of this thing. We shall see.

http://www.alphacool.de/images/produ...es/12273_1.jpg

Link

For air cooling the PWM area, you may want to track down one of these fans that ASUS and ABIT provided as an option on some of their boards to cool the heat-pipes...

http://techreport.com/reviews/2005q4...sli/blower.jpg

You might have to do a bit of modding and/or use superclue to get it to stay on the DFI heat sink, but it should work.

Alternatively, you could cut a hole in the large blank area of the I/O plate adjacent to the PWM area and mount a 40mm fan to the exterior so it's blowing air through the hole in the I/O plate over the PWM area.

Quote:

---

Originally Posted by OldGuy

Everest also reports the core temp and it is spot on with coretemp...

---

There ya go, you learn something new everyday. :)

VirtualRain - Your work with this board is fantastic. Well done.

Quote:

---

Originally Posted by virtualrain

Here's what I've figured out about monitoring on this board...

BIOS, Speedfan, and SmartGaurdian all pull monitoring values from the ITE SuperIO Chip the IT8712F which includes voltages and a few temps (CPU, NB, and SB).

Note for those who don't already know... the temp in BIOS called "PWM Area" is mislabled... this is actually the NB temp.

---

I've been working on the assumption up to this point that this also means that the temp reported as "System" by Smart Guardian is actually the NB temp since this category would normally refer to the PWM Area and that the "Chipset" temp reported by SG is actually the PWM Area. Does this seem to be the case with you as well?

Quote:

---

Originally Posted by virtualrain

Wow... that seems way high... Perhaps try remounting the NB sink.

I just have a basic 120mm fan leaning up against the NB cooler. I guess it's pretty effective since it cools the top and base together. I'm working with it open on the bench though so that helps as well. I'll be water cooling it when it eventually goes in the case.

Here's pics... http://virtualrain.blogspot.com/2007...-dfi-680i.html

---

And the vid card with that fan sitting like that:confused:

Quote:

---

Originally Posted by century child

I've been working on the assumption up to this point that this also means that the temp reported as "System" by Smart Guardian is actually the NB temp since this category would normally refer to the PWM Area and that the "Chipset" temp reported by SG is actually the PWM Area. Does this seem to be the case with you as well?

---

Not exactly... in SmartGuardian... "System" = SB, "Chipset" = NB.

Quote:

---

Originally Posted by palese

And the vid card with that fan sitting like that:confused:

---

LOL! The fan is not touching the video card but even if it was, it's not a problem (a) because the fan is plastic and (b) the video card is a POS! :D

I'm just using a POS MX4000 PCI video card for the bench setup since I want to keep my old system up and running until I'm satisfied with the OC and ready to mount it all in the case. I work at home so I can't afford any down time and it's also why I'm a bit more of a prime-stable fanatic I guess.

Thats it, I'm pulling it all apart tomorrow night, cleaning all the surfaces and trying some different paste.

NB temp was at 50c for 3 hrs of prime, then shot up to 70c and seems that the NB HS is not completely stable on the chipset, if i push on it a lil the temp drops, I would say its a bad mount, how about that foam tape on the board and HS, is it necessary to have on there for stabillity? I'm thinking about taking it off for a tighter seat.

Quote:

---

Originally Posted by xTr33m sammich!

I'm thinking about taking it off for a tighter seat.

---

there's a thread about this on dficlub...

http://csd.dficlub.org/forum/showthread.php?t=2650

I've a solution for the PWMs:

Wait about a few days, MIPS-Computers is producing new PWM-Coolers for the DFI board!

They design a new NB Cooler, too. The Intel Freezer for the Bad Axe Board does not fit well.

Fixed for now, dang it...
NB under load at same setings is now at 46c
flipped the NB HS so the lanparty logo is upright, lost an inch or so clearance for ultra-120 ex so I had to flip it 90degree too, post a pic tonight, man it's tight...

the pwm mosfets are very small, they are similar to the ones found on the ATI 1900 cards at first approximation. These are delicate devices, and do not take well to skewed pressure on the die, if you are going to fiddle with the stock heatsink do so under caution and use a delicate touch at all times..

regards
Raja

Quote:

---

Originally Posted by xTr33m sammich!

for ultra-120 ex so I had to flip it 90degree too, post a pic tonight, man it's tight...

---

ya please do i'd like to see it...

on another note regarding mips computers please let us know when this gear becomes available...

After tearing it apart, flipping the s#!t, and prying it back into place temps are fairly good, I wanted the Ultra120ex to be mounted vertically exhausting out the rear of the case, instead, since my NB overheats with its HS mounted upside down(doesnt dfi picture it upside down?) this is the only way it all fits together, all fans blowing up.. btw, the fans are what I had laying around from previous builds, the NB fan is a zalman 80mm fan (spare from the GPU HSF) it works, whatever, it's my not so quiet xTr33m sammich!

http://a423.ac-images.myspacecdn.com...a79c9a4386.jpg http://a983.ac-images.myspacecdn.com...5919fab7be.jpg

http://a497.ac-images.myspacecdn.com...8670e25758.jpg

Quote:

---

Originally Posted by xTr33m sammich!

After tearing it apart, flipping the s#!t, and prying it back into place

---

uhmm wow ! great pics. couple questions... are you concerned about weight with the two 120mm on the ultra ? was it an issue at all like did you hear the motherboard creaking lol... is the top fan on the ultra clipped or is it just laying on it.

when you say flip the NB upside down you mean rotating it 180 degrees from where it is now ? good to know there could be overheating issues with it like that...

looks like you aren't using a back case exhaust fan?? is that to do with negative airflow issues or just cause you ran out of fans lol