• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

*** Official NVidia nForce 680i Thread ***

Page 6 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Thank you for the screenshot, Madellga. Now I have to figure out what I'm doing wrong with chipset/memory, I guess. My Xeon passes Small FFT fine but fails Large FFT anywhere between 1~3 hours so I was sorta suspecting chipset overheat. I will try less voltages for SPP.

P.S. Well, at least E6400 runs cooler than E6600. Even at 1.65V temps never rise above 70C. 😉 But yeah it pisses me off. I mean, the thing does 3.20GHz at default voltage, then it doesn't even boot @any frequency between 3.20GHz~3.76GHz. LOL.
 
I think it is a common problem. I can get mine to run also at 3.6GHz, it can run Orthos for some hours but it fails between 6 and 12, sometimes less.
I can bench it at 3.6GHZ, but not running 24/7.

3.4 is rock stable, 3.5 is doable but VCore goes up. As for the E6400, mine had a wall also at 3.2GHz, but recently the wall was pushed to 3.4 - thanks to Corsair ram?

I agree with you, once the bios bugs are sorted out, this is a better platform than the 975 - even if Superpi is not faster. The difference is too small in real life.
 
The unstability under Large FFT run was caused by insufficient VDIMM/VSPP. Up'ed VSPP by 0.05 and VDIMM by 0.025, and now it runs fine. (At least for 4 hours until I stopped) One word of caution: As I stated somewhere in this thread, the board overvolts VDIMM. (measured by DMM) The actual voltage for memory will be 0.04~0.06 higher than what you see in the BIOS. I haven't bothered to measure VSPP, but on some sunny day I will find it out and report back. 😀

Originally posted by: Galadul
Hey, I have the same system, e6600 and evga 680i, corsair dominator 2gb 4-4-4-12 (the ones with DHX), BigTyp 120VX HSF

I had two questions...
1. First, my Hiper 580w PSU, which I thought was pretty kick ass, and certainly was on my AMD system, has a 4 pin ATX connector, not an 8 pin. Is this a real issue? Should I not be overclocking until I get a 8 pin atx PSU? For now I'm running 3600 @ 1600 fsb, mem 1:1 synch. Seems stable, but haven't had time to run any comprehensive stress tests, and this weekend I'm whisked away for work. (Any free stress test programs to recommend?)

2. My ram, Corsair Dominator TWIN2X6400C4D 2048MB. Anyone have any experience overclocking these babies? Currently running them at 4-4-4-12 2T @ 2,1 volt as per factory specs. Now, I'm only interested in gaming performance, not how long it takes to convert wav to mp3 or various memory benchmarks or any of that. So basicly I guess what I'm asking is if there's any point to raising Mhz or lower the timings on these babies if I'm mainly concerned about gaming?
Click to expand...
1. Missing 4 pins shouldn't be a problem as long as your PSU is a quality one. It's a sticky issue and you're the only one can truly determine whether it's stable. If you run enough torture tests and the system takes it without resistance, then everything is good. I used to run DFI NF4 SLI-D board with no less than 10 x A64/Opteron CPUs, with 20-pin main power connector plugged in. (The board has a 24-pin power socket) Stability was top notch and Vcore never dropped/fluctuated like in Intel boards. I also achieved record OCs at the time. (Opteron 165 @3.0GHz)

2. I would say no. Although some folks like to play with memory frequency/timings (including myself), it's generally not worth it. (I'm being honest) There is little to no performance gain and if there is, it'll only show in benchmarks. Unless Sandra is your main app, you should be glad your memory isn't causing you grief and enjoy your games and music! (There are many folks dealing with memory compatibility issue on this board)
 
Thanks alot lopri 🙂 Warms my heart to know you ran a 24 pin mobo with 20 pins 🙂
I haven't actually done any overclocking in like 4 rigs, which is...uh...alot of years ago and more CPU/chipset generations ago than I care to count. I think it was in the legendary Voodoo2 days. Having this overclock friendly rig and NOT overclock would be criminally insane however.

About X-fi: I'm using X-fi Fatal1ty w/ front IO panel and have NO issues. (if you change windows speaker sync, you need to restart or you will get popping sounds, but I had this issue with AMD mobo too, and anyway the creative panel tells you to reboot.)


I think I used your bios images lopri in another thread to overclock, which means I'm running cpu v 1,49275 (bios reports 1,44 I think, Ntune 1,50something)
cpu fsb 1,4
SPP 1,45
MCP 1,525
HT SPP - MCP 1,25

Now, you've been playing around alot more than me with this board/cpu, and probably some more after you posted those bios images. Any voltages I should adjust? I saw you write something about SPP and MCP voltages, but I don't know if you only needed to adjust those to run @ 3.7 mhz. Either way, if 3.6 is difficult to keep stable, it prolly won't kill me to do 3,4 @ 1500 FSB or something. But then I should prolly lower CPU voltage...hmm..this OC business sure takes time 🙂

Edit:
Oh and another thing, late last night I experienced some graphic troubles. First in windows, but was fine after a reboot, but then had some discolouring in a 3D game. I didnt have time to test anything else or play with voltages though. My GPUs aren't overclocked, they're both getting power from separate rails, and the mobo EX is plugged in and enabled. I have, however, set PCI x16 lanes at 102 instead of 100...could this be the issue? And in general, can CPU/FSB overclock cause graphical errors? BTW, bios is P20 and Nvidia graphics drivers is...uhm..93.71? It's the newest non-beta drivers anyway. I'm @ work right now so I can't remember all this in my head lol. (Yeah I have crazy work hours, and the thought of spending 10 hrs watching orthos drives me insane)
 
Well...changed PCI x16 to 100, ran Orthos for the first time, using large FFT, PC lock up after 9 minutes..now that can't be good 😉
I guess upping CPU core is next

Edit:

Well, upped cpu v to 1,5 which is read in bios as 1,45. Orthos Large failed after 18 min. Progress!

At this time ntune reported CPU temp at 71. I rebooted rather quickly, and went directly into bios, and bios said cpu temp was 39....so

1. TT BigTyp 120VX cools the CPU faster than you can say "Warranty denied due to OC"
2. Bios cpu temp reporting is wildly innacurate
3. Ntune cpu temp reporting is wildly innacurate.

IF the temp was actually 71, I don't think I should up the cpu v anymore..don't you think?
That's all the OC messing around I have time for today... it's 10:20 PM here, and my bottle of vodka is calling for me.
 
I also raised Vcore after finding out my system was not totally stable. Since I knew my CPUs capability before getting this board, I considered 20~30 mins Orthos as good enough. Later I found out the difference between actual Vcore supplied and BIOS reading, and Orthos wasn't stable for longer period of time. Vcore fluctuation wasn't too bad, however. Playing with the right ammount of voltages is critical for OC'ing on this board, IMO.

CPU temperature rises proportionally to frequencies. Giving higher voltages will of course contribute to temp rise, but not as much as higher clocks. You will notice higher temperatures with higher Vcore when CPU is idle more easily. Under load, the difference is less pronounced. (comparatively) 71C is somewhat high. Not because the chip will die soon (it will still live long enough), but because it will affect stability. On the same line, I would select up to 1.55V on this board with 4MB L2 Conroes, only if I could manage to keep the temperature under 65C. Past 65C, CPU will show symptoms of instability at any frequency. nTune's temp reading seems reliable on this board so I would go by it. I didn't really pay attention to BIOS temp reading because I think what matters is load temperature. Try re-seating the BT? Also make sure the case has proper ventilation for the blowed-down air by BT. Without proper ventilation the hot air stays around CPU!

For finding out optimal Vcore, use small FFT first. Once you find it, Vcore for large FFT will be close. Large FFT stresses all of CPU/NB/memory and is my personal favorite for total system stability. Finding optimal NB (SPP) voltage can be tricky with this board, and worse yet, the SPP gets hot very quickly, so again the need for proper ventilation. From the looks of it, if you can keep the CPU temps down to <65C, your CPU stability is pretty much guaranteed (test with Small FFT).

My current BIOS configuration @3.6GHz/DDR2-800 (3-3-3-2T)

VCore: 1.51875V
VDIMM: 2.2250V
VSPP: 1.45V (Try 1.30~1.50V)
VMCP: 1.55V
VSPP-MCP: 1.30V

And my readings go by:

VCore: Minus 0.04~0.05V from what's set in the BIOS
VDIMM: Plus 0.05~0.06V from what's set in the BIOS
CPU Temp: nTune
 
yeah makes sense what you're saying. My case IS kinda too small for my setup. the BigTyp has 3-4 mm clearance to the underside of my PSU, but since the PSU has 120mm intake fan on the underside, I thought it might work out. But since temps are so high, it's probably not ideal. I guess I should get some ugly stacker tower or something...
Also, the side door 120mm intake fan on the NZXT Lexa is really far down, meaning it's blowing directly onto the GPUs, not the CPU. 71 was measured with the side door off though, so the case was open and the BigTyp should have gotten all the air it needed. No cables or anything around the BigTyp to hinder airflow, but yeah that PSU is mighty close.

Anyway, what you're saying is I should test with orthos small, and that for orthos large, 30 min might be good enough?
 
A word of caution (read below)

I finished last night a 24 hour Orthos run (BLEND) to verify my settings, which I consider final and 24/7 safe and stable.
http://img206.imageshack.us/img206/1603/680ie6600vcore140vdim21hl8.jpg

The computer was shut down through Win. This morning, it would not boot: I turned on and 2 seconds later it would shut down.

Before panicking, I tried to reset the bios, no effect. I swapped components with my rig 2 (video, memory), no change.

Last, before sacking the CPU off the board (and wasting half a day on rebuilding the system), I tried removing the CMOS battery.
It did the trick!!

So if any of you face a similar situation, turn off the computer completely, unplug from mains, reset the BIOS and remove the battery.
That can save you from a lot of unnecessary work.

As for why it happened, I have no clue. Anyone has some ideas?

As for BIOS settings, my Voltage settings are:
VCore: 1.45V
VDIMM: 2.10V
CPU FSB: 1.3V
nForce SPP: 1.3V
VMCP: 1.50V
VSPP-MCP: 1.20V

I'm just experimenting and going down with VDim step by step. FSB and SPP must stay at 1.3V, otherwise machine will hang up sometimes. +0.1V MCH/FSB is the same increase the Intel chipsets needed.
 
Galadul:

What Orthos Small FFT does is isolating CPU and testing whether its capable of your OC. It stresses Floating Point unit and L2 cache of the CPU. Large FFT tests CPU+NB+Memory. Blend also tests CPU+NB+Memory but it tests lots of memory (more than 60% of total system memory) so often times you won't be able to use your comp while its running. Large FFT is very intense, and I find it perfect tool to determin NB stability.

So for instance, if your system passes Small FFT but not Large FFT, you can be relatively sure that it's not the CPU causing it, but rather NB or memory. Blend test is also excellent, probably closer scenario to actual gaming usage, but I prefer Large FFT because Blend requires more time, and as mentioned above often times the system becomes not usable. But this is just a personal choice. Either way, 30 min pass is nowhere near the ideal stability. What I said is that for initial testing, I just did 20~30 min quick tests since that I know my CPU can do 3.60GHz around 1.50V. After you find your optimal configuration, you need at least 4 hours (preferably 8 hours) of Large FFT or Blend.

Madellga: I wouldn't worry about it unless you encounter similar problems again. Man.. 24 hours Blend? I would consider that mega ultra super duper turbo belly stable. Knowing this board does weird things, I would wait till next BIOS update until I make any judgement. I'm also thinking (and hoping) the weird behaviors are due to sh*tty BIOS, not because of hardware.
 
Has anyone attempted lifting up the heatpipe/heatsink and re-applying Arctic Silver? I'm seriously considering it but would like to know what the factory applied TIM looks like.
 
Originally posted by: lopri
Has anyone attempted lifting up the heatpipe/heatsink and re-applying Arctic Silver? I'm seriously considering it but would like to know what the factory applied TIM looks like.
Click to expand...



I replaced the stock goop with arctic silver. I never tried running it with the stock stuff, so I don't know if it helped or not. If it did, I'm guessing not by more than a couple degrees based off what I've seen reported by others.

I found the mcp fan to be too loud even at 50% so I have a 80mm Nexus fan wedged between the heatsink and my video card, a 7600gt silentpipe. It idles at around 60C or so I think.

Which brings up a question. Does anybody know of any tools to monitor the mcp temperature from Windows? I'd like to know how hot it is getting under load, but I haven't found a way to monitor this yet.
 
yeah i did the same thing, got rig of that crap goop stuff and put AS5. but i didn't test the stock goop temps first either. i did like how easy removing the heatpipes and heatsinks was. i have done that to a few mobos in my day and this mobo was by far the easiest of them all.
 
It's been several weeks since the striker was introduced. Yet still, no one has them in stock. I beginning to think Asus is just showing pretty pictures and really doesn't have anything ready. Maybe all the other 680i boards are just a myth and the eVGA is the only one we'll have. I haven't heard a peep about the DFI 680i in months either. Just release the damn board already so I can play at a resolution higher than 1024x768. 🙁
 
The Striker is out. Limited quantities. A lucky few got them. Basically a paper launch so far. Now that Turkey Day is over should see some more deliveries. Hopefully.


Edit: Pricegrabber is showing it in stock at Ajump and ATACOM at the moment.
 
Originally posted by: Skott
How much change in temps, if any, did you get, Lopri?
Click to expand...

Hard to say. There is an item called 'System' in nTune temperature monitoring and the temp indeed went down from 36C to 31C, but I have no idea what it's referring to. It could be NB, SB, or somewhere on the motherboard.
 
WARNING 1: The discrepency between actual voltages supplied and what's set in the BIOS is bigger than I thought them to be. It's so disappointing that NV have to 'lie' about real voltages and/or they're incapable of providing correct readings/knowledge in the BIOS. Whatever the case, please note the following and be extra careful to keep your expensive gear in good shape. (If possible, measure your own)

vCore: The board UNDERVOLTS by 0.04~0.06V
vSPP: The board OVERVOLTS 0.07~0.08V except @1.55V
vDIMM: The board OVERVOLTS by 0.06~0.07V up to 2.30V, and past 2.30V it OVERVOLTS by 0.09~0.10V

I was clocking up my memory and noticed weird behavior of the board. I was able to push memory frequency linearly up to vSPP 1.50V. (Of course vDIMM was raised gradually, too.) But as soon as I set vSPP 1.55V in the BIOS, the system became unstable. So I picked up my DMM and measured the vSPP and compared it with the BIOS reading. The results;

BIOS -- Actual
1.30 ---- 1.38
1.40 ---- 1.48
1.50 ---- 1.58
1.55 ---- 1.46

As you can see, the board overvolts SPP massively, up to 1.50V, then undervolts, again massively, at 1.55V. vDIMM was another surprise. I up'ed the vDIMM to 2.325V and DMM read 2.420V. I know many reviews don't measure accurate voltages for individual ICs, but I would think such big discrepancies could have been discovered by someone. 😕

Warning 2: Whether you OC or not, always use the supplied fan for the board's heat management. Without the fan, SB temp rose up to eye-popping 80C. (Can't keep my finger on more than 2 secs) With the fan, MCP generally stayed under 50C, with occasional spikes up to 55C.
 
5C sounds about right. Thats what people saw on the 965s and 570s mobos on average. Yeah, all 680i mobos so far have a big vdroop from what I been hearing.
 
This is a double post from the evga 680i topic ... apologies ... but ...

All? Are the EVGA and the BFG and the Biostar 680i identical to the point where one BIOS will work on any of the others?

 
I've just acquired 2 1Gb sticks of OEM Micron D9 that defaults at DDR2-800 5-4-4 at 1.8V. I'm wondering if those sticks will work with a 680i mobo. I'm either gonna get the Asus or EVGA model. But before I do I want to make sure these sticks will work. Thanx 🙂
 
Originally posted by: trake1
This is a double post from the evga 680i topic ... apologies ... but ...

All? Are the EVGA and the BFG and the Biostar 680i identical to the point where one BIOS will work on any of the others?
Click to expand...
Yes.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top