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

[Jobeo]

1.Fixed Core Multi-Processing function fail issue on Intel Core 2 Extreme QX6700 /X6800 CPUs.

what exactly does that mean ?

[virtualrain]

I suspect it's a bad translation and should have read... "Fixed: multi can't be set above x10 on extreme processors"

Just a guess since this was reported as an issue in the previous BIOS (Link). It would really be helpful if DFI could be more explicit in the change log... surely this isn't the only change.

[Jobeo]

well winflash went fine... although it did stop at 12% and just sat there. thank god i let it finish cause it finally came up and said it was a success.

so far i can see a couple changes. the cpu shutdown temp defaults to disabled. and strangely it is underclocking my CPU by defaulting to an 8 multi.... with the last bios it correctly defaulted to a 10 multi but now it starts with an 8.

GTL values still seem mislabeled. i will clock around with it a bit and see if its any easier.

[Jobeo]

ahh... they finally fixed the temp labeling in the BIOS... NB and SB are now properly labeled... no more PWM stuff

[Borgod]

Max multiplier still only 10?

After playing with my DFI NF680i for only a short period of time it ran my QX6700 at 3400Mhz prime stable for 8 hours @ 1.45vcore

At 3500Mhz overnight I woke up to see the PC was off, so either it reached its max temp of 90degrees (which I doubt... it's sits around 80 degrees all other times while priming at this voltage) or it crashed and didn't boot up.
CoreTemp log tells me it didnt' reach sky-high temps either, taking into account the 15degree difference between CoreTemp and SmartGuardian (BIOS). I think it crashed at about the 2.5hour mark which is a lot longer than I thought it would last tbh!

I'll push for 3550Mhz tonight and see if I can get it to crash sooner, so I can narrow down where I need to focus the tweaks.
I also need to lap my QX6700.

Cooling specs:

DtEk Fuzion,
PA 120.2
DDC+ /w Petras top

[vincey2kr1]

What powersupply are you using? I get the same type of crash when I try at 3.5ghz but I'm only aircooling, still playing around at 3.4ghz 1.45v on q6600 Enzotech ultra120

[Borgod]

Enermax Galaxy 1000w

[vincey2kr1]

Eek last person who was running a Galaxy with this motherboard was having the same problems at Stock speeds.. I get the shutoff as well with my Thermaltake 1000w havent really tried stock speeds though.

[Borgod]

Hmm I've just sold my Silverstone Zeus 750w

I wouldn't have thought the shutdown was related to PSU but?

[virtualrain]

Hmm I've just sold my Silverstone Zeus 750w

I wouldn't have thought the shutdown was related to PSU but?

Search back in this thread for Nugzo's experience with the Enermax Galaxy... His problem was more severe so it might not be the same issue... your problem might not even be PSU related but if you find it's the same it would be the second such issue with this PSU and board combo.

[Jobeo]

I wouldn't have thought the shutdown was related to PSU but?

what kinda cooling do you have on the PWM

[Borgod]

I showed you at DFI Club, I'm Terebro

virtualrain: Cheers for that. Indeed my symptoms aren't anywhere near as bad as his. It did run prime Stable for 8+hours at 3400Mhz and it's ran flawlessly otherwise including hours worth of gaming.

However I'll continue to have a look around.

I've still got a long way to go, I've only set the GTL's once and want to rotate my NB 180 degrees aswell. Not to mention install XP for more serious benchmarking.

On a happier note, in hwbot my comparable 1024 and 32M SuperPi speeds are the fastest at whatever frequency I set my QX6700 to It's even faster than several that run a higher frequency

[Borgod]

Jobeo: Did you see if flashing BIOS retains the 4 'saved' user-defined BIOS settings? Or do we lose them?

[Jobeo]

Jobeo: Did you see if flashing BIOS retains the 4 'saved' user-defined BIOS settings? Or do we lose them?

it keeps them...

[C-N]

At 3500Mhz overnight I woke up to see the PC was off, so either it reached its max temp of 90degrees (which I doubt... it's sits around 80 degrees all other times while priming at this voltage) or it crashed and didn't boot up.
CoreTemp log tells me it didnt' reach sky-high temps either, taking into account the 15degree difference between CoreTemp and SmartGuardian (BIOS). I think it crashed at about the 2.5hour mark which is a lot longer than I thought it would last tbh!

I'll push for 3550Mhz tonight and see if I can get it to crash sooner

I get a similar thing with my E6400.

I can run 450 x 8 3.6GHz @ 1.55v Orthos stable 8 hours + but if I try 462.5 x 8 3.7GHz it will fail Orthos blend quite quickly. Orthos takes longer to fail the more voltage I add until I get to about 1.6 + 1.55% = 1.625v then it will run for 30-45minutes & then just shutdown. It always fails on the CPU test 8-10 or 12fft.

The funny thing is I can run small fft's for hours @ 1.625v but it wont make an hour on blend.

Core temp never goes above 71oC which is strange as you would expect the CPU to make more heat the more voltage you add which leads me to believe Core temp is not reading correctly & maybe the CPU is hitting its shutdown temp.

I also notice my Smartguardian temps steadily rising just a few degrees off Coretemp until you get over about 1.5v then Smartguardian keeps going making about 10-15oC difference up at 1.6v.

What I am trying to say is Smartguardian apears to increase steadily just a few degrees 3-6oC above Core temp until a certain voltage then Coretemp stops going up & Smartguardian keeps going eventually giving you the 15oC difference under load.

I am not suggesting my CPU can do more than 3.6GHz but on all my other boards the CPU would either fail Orthos or reboot the system rather than just shutdown which leads me to believe this thing/problem is a temp or reported temp shutdown issue.

The shutdown still happens on the latest BETA BIOS (521) & the shutdown temp is disabled in BIOS even though none of the temp reporting programs get to the max shutdown temp.

It would be interesting to see what you make of it when you have a play tonight.

Regards

Craig.

[Jobeo]

The shutdown still happens on the latest BETA BIOS (521) & the shutdown temp is disabled in BIOS even though none of the temp reporting programs get to the max shutdown temp.

its a known issue that if the PWM overheats the rig will shutdown unexpectedly so my guess is the PWM getting too hot. unfortunately we don't seem to have a temp sensor for the PWM so there's no way to know when this will occur.

to test it i'd put a 120mm high-speed (~2000 RPM) fan blowing as best you can on the PWM and run your tests to see if the shutdown still occurs. at least then you'll know if this is whats causing it.... although mounting such a fan for long-term use would seem impossible with the layout of the board =/

[C-N]

its a known issue that if the PWM overheats the rig will shutdown unexpectedly so my guess is the PWM getting too hot. unfortunately we don't seem to have a temp sensor for the PWM so there's no way to know when this will occur.

to test it i'd put a 120mm high-speed (~2000 RPM) fan blowing as best you can on the PWM and run your tests to see if the shutdown still occurs. at least then you'll know if this is whats causing it.... although mounting such a fan for long-term use would seem impossible with the layout of the board =/

I have measured the load temp of the PWM with an infrared/laser thermometer & it measured mid 50's under load. I have also put a 70mm fan over the PWM heatsink & got the temps down to low 30' & it still shuts down.

Regards

Craig

[virtualrain]

The funny thing is I can run small fft's for hours @ 1.625v but it wont make an hour on blend.

This to me means it's NB or RAM causing you problems. Small FFT's fit in the CPU cache. If you can run small FFT's for hours then your CPU is stable. If it fails blend then your NB or RAM are causing errors. Jack your Volts on those two and see what happens or loosen your timings.

As for the shutdown, it could be a few things as mentioned 1) PWM overheating, 2) Thermal shutdown, 3) Hard-core failure. If you are already cooling the PWM area, then my only suggestions are: rather than disable the thermal shutdown, try setting the thermal shutdown to the highest threshold setting. Perhaps there is a bug with disabling it. Perhaps try the new beta bios. Finally, try backing off your overclock somewhat and making sure you can get something stable that doesn't cause a shutdown.

[C-N]

This to me means it's NB or RAM causing you problems. Small FFT's fit in the CPU cache. If you can run small FFT's for hours then your CPU is stable. If it fails blend then your NB or RAM are failing. Jack your Volts on those two and see what happens.

Thats what I thought & I did gradually push my NB up from 1.51 to 1.64v & it made no difference.

I also unlinked my RAM & left it at DDR900 where it had run Orthos 8 hours + & later I gradually bumped the vdimm up from 2.18 to 2.38 in BIOS & it made no difference either.

I will have another play later & see how long it will run small fft's.

As for the shutdown, it could be a few things as mentioned 1) PWM overheating, 2) Thermal shutdown, 3) Hard-core failure. If you are already cooling the PWM area, then my only suggestions are: rather than disable the thermal shutdown, try setting the thermal shutdown to the highest threshold setting. Perhaps there is a bug with disabling it. Perhaps try the new beta bios. Finally, try backing off your overclock somewhat and making sure you can get something stable that doesn't cause a shutdown.

It will run 450 x 8 3.6GHz fine & I am happy at that its just when I was messing around the other day I noticed this shutdown thing & thought I would post my findings.

My thermal shutdown is normally set to 85oC I only disabled it to try to rule the thermal shutdown out which it didn't. I am currently on the new BIOS & its just the same.

Regards

Craig.

[virtualrain]

Hey Craig, it sounds like you were several steps ahead of me then... I wish we could get to the bottom of this shutdown issue.

In other news, there is a new thread in the reviews section on this board...
http://www.xtremesystems.org/forums/...d.php?t=146067

There's also a BIOS mod community playing with this board called "Rebel's Haven" (never heard of it)... The seem to think we all have cherry picked parts over here! (Couldn't it just be we are good overclockers?! )

http://www.rebelshavenforum.com/sisu...ubb=forum;f=54

[Raja@ASUS]

If the board bows slightly there is a chance that the central mosfets fail to make decent contact with the PWM sink.

One thing you can try is to use some extra Ceramique under the PWM sink to ensure some form of contact at all times, make sure you use non conductive/non capacitive thermal paste. Ths may help the shutdown somewhat, juction temp at the mosfet will be substantially higher than the 50 degrees you see on the heatsink.

Make sure you are careful if you decide to remove the PWM sink, do not skew the heatsink with excessive pressure..

regards
Raja

[C-N]

Raja I will get some Ceramique & give that a go on the weekend along with ensuring the heatsink is flat.

The rig is running Orthos small fft's at the moment (2 1/2 hours so far) at 462.5 x 8 3.7GHz with a 70mm fan on the PWM heatsink & a 120mm blowing down between the CPU & NB toward the long chips in front of the PWM heatsink. The heatsink temp is mid 20's & the long chips are low 30's.

One thing I now notice is Smartguardian is running very close to Coretemp (high 60's low 70's) maybe its the fans maybe its the later 521 BIOS I will investigate later & post my findings.

Regards

Craig.

[owcraftsman]

I have always heard and read explanations that were way above my head that the 1/2 (.5) multipliers should not be used. Try a slight step up.
Best of luck

[Grinch]

I have always heard and read explanations that were way above my head that the 1/2 (.5) multipliers should not be used. Try a slight step up.
Best of luck

I have heard that as well....

[virtualrain]

I don't think anyone is using a half multipler here. That's not even an option with Intel.