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

[trake1]

Thank you! Yes the EVGA P21 BIOS installed perfectly on my BFG 680i, and the system is running perfectly also. Used the nTune BIOS updater downloaded from evga and followed directions exactly.

 

[blank98]

Ok I got this motherboard and im getting 3 different temp readings, my mobo is telling me that my CPU is 34C idle then when im in windows TAT is telling me 50C idle and nVidia monitor is telling me 44C idle so which one is right.

 

[lopri]

blank98, I find the nTune (nVidia monitor) and TAT (CoreTemp) report similar temperature on my board. There is normally 2~3C gap but it's been consistent. I would first disregard BIOS reading of CPU temperature. Doesn't TAT report each core's temperature separately? How far apart the TAT temps is from nVidia monitor temps?

P.S. I've taken some BIOS shots to prove my point regarding the SB temps and the 'optional' fan. As you can easily guess all these temps are 'idle' temperatures, since it's not possible to 'load' anything while in BIOS. The fan MUST be used under any configuration, IMO.

With Fan: http://img50.imageshack.us/img50/3444/dsc01191pn7.png
Without fan 1 mins later: http://img50.imageshack.us/img50/8969/dsc01193hm9.png
Without fan 2 mins later: http://img50.imageshack.us/img50/8969/dsc01193hm9.png
Without fan 3 mins later: http://img50.imageshack.us/img50/9681/dsc01196em0.png
Without fan 4 mins later: http://img50.imageshack.us/img50/7789/dsc01200nr2.png
Without fan 5 mins later:http://img129.imageshack.us/img129/6151/dsc01205lm4.png
Without fan 10 mins later: http://img129.imageshack.us/img129/4544/dsc01209iz3.png

 

[blank98]

CPU 0 is usually at 50C and CPU 1 is at 48, for some reason cpu 1 is always cooler than cpu 0.

 

[lopri]

Are CoreTemp and nTune temperature always far apart? The difference between CPU 0 and CPU 1 is normal as long as they stay in 3~4C. If nTune and CoreTemp readings's difference is within 5C, just go by either one. I know even a few degree of difference could be sometimes uncomfortable, but with not so many software readings available for this board, that's the best way to deal with it, I think.

Another note regarding my previous post: I now replaced the stock HSF for NB/SB with thermalright HR-05-SLI, and I am positive to say the incredibly high MCP temperatures in BIOS is, actually, the temperatures of SPP (Northbridge). Before when NB and SB are connected via heatpipe, I was mistaken the high temperature was that of MCP's due to the BIOS reading. Separated, MCP remains cool but SPP makes the Thermalright heatsink blazing hot.

 

[skrewler2]

HR-05-SLI fits fine with the 8800GTX and .. what CPU HSF are you using?

 

[lopri]

HR-05-SLI will fit fine with most aftermarket HSFs. (Especially for SPP) I use Scythe Infinity and there is ample room between that and the 8800 GTX.

 

[lopri]

DDR2-1266 in Action.

http://img182.imageshack.us/img182/6103/ddr1266pzd4.png

 

[skrewler2]

Makes me almost regret going with a 6600 and not a 6400, 6400's seem to overclock so much better

 

[Skott]

Dont discount the E6600 out of hand. Over on XS some are using it with good overclocks. One has over 500fsb stable with a E6600 combined with the Striker Extreme and its a 680i mobo. Granted its different than the reference boards eVGA, BFG, and ECS put out but the potential is there.

 

[lopri]

I think I was wrong again with the temperatures.

The MCP temperature reading in the BIOS is correct. (which gets insanely hot without fan working) My guess is there is no SPP temperature reading available in the BIOS nor in nTune, and for a good reason. NV probably couldn't disclose the actually SPP temperature because it's just unreal.

 

[lopri]

I'm also having what looks to be memory controller and/or disk controller related problems. At first I thought it's the NV IDE drivers and/or BIOS issue, but counter-proofs are popping out in an unpredictable fashion. Chipset-related problems are hard to explain and I don't have a solid proof so far, but I'll have to track it down, too. At first I thought it's heat-related, but those weird issues haven't disappeared even after I managed to cool the SPP/MCP down. (Also one of my PCI slot seems to be dead. Ugh..) Granted the way I handle my hardware can only be described as a 'torture', but the board shouldn't crap out in a week no matter what.

I'd I think I've spent way more time than I hoped on this board, and really want to finish this build real soon. I will keep you guys posted.

 

[DanBaer]

Originally posted by: Accreta
I don't understand why anyone would spend so much money to purchase a Striker?!? Compare the P5N32-E SLI and Striker. The only big difference is the LED read-out, and is it really worth an extra $100? I like Asus but not that much. I'd choose the P5N32-E SLI anyday...

Compare features HERE

Price comparison of P5N32-E SLI and Striker

Those solid capacitors are worth the difference by themselvs. I will bet thet they will extend the life of that board by at least 50%.

 

[Skott]

Even though the Striker Extreme has more features in the end its going to have to overclock considerably more than the P5N32-E SLI or otherwise most people will just opt for the P5N32-E SLI and get just as good of a overclocking mobo. Some are already calling the P5N version the, "Striker-Lite" mobo.

 

[politicalslug]

I love this motherboard. I have gone from a Gigabyte DS3 to an Intel Bad Axe to an Abit AB9-pro to an Asus P5B to a Gigabyte DQ6, then finally settled on this Evga 680i. This board has everything I need (except for two additional sata ports, but i fixed that with a pciE sata card). I have OCZ 800 platinum cas 5 (but it uses D9s) and I've pushed the memory to 1120 @ 4 4 4 12 2T @ 2.325v. CPU is running at 3.15 and 1.4v (less than stock on this board, which is 1.46) and everything is totally stable. Prime95 for 10 hours and totally stable. CPU temps at load are 59-61c, but still totally stable. I haven't tried a higher overclock yet, but i feel confident i can take it higher. I can confirm that, at least in my case, this board overclocks more easily and will far fewer issues than any of the above mentioned boards.

Anyone else have experiences with board and want to share with the community?

 

[lopri]

I've resolved the memory issue. Turned out to be my misunderstanding.

But one of the PCI slots is still not functioning. Not sure how that happened because it worked once before, and hadn't touched till yesterday. It's not a deal breaker (at least I have the other PCI slot working) but certainly is unpleasant. Will see if it comes back to life. But other than that, everything has been worked out.

After all these frustrating/rewarding moments, I finally settled at following settings:

CPU: 3.60GHz (9x400) / 1.51875V (est. 1.47~1.48V)
Memory: 1100MHz|4-4-3-9-2T (Unlinked) / 2.325V (est. 2.40~2.42V)

Only other thing I might tinker with is getting 1T to work when a decent BIOS comes out.

Overall I don't regret the switch from 975X and the board is overall very solid. It may not be as snappy as 975X, but the build quality seems higher than that of P5W-DH, its layout is near perfect, very versatile and feature-packed (sans eSATA) and it gives me the possibility of SLI in the future, so I'm satisfied. And I know I can rely on EVGA's RMA. (Advanced RMA rocks!) Can't say the same for ASUS.

If I were to give advice for other users: Everyone will have moments while trying to squeeze more performance out of this board, but in any case, KEEP THIS BOARD COOL. This board is one of the hottest boards I've ever dealt with. (No wonder ASUS had to resort to such bizzare heatpipe mess) Not only the SPP/MCB, but also other spots including VRM. Erroneous voltage/temp readings in the BIOS don't help either. Keep good airflow inside the case and, if possible, employ active cooling on board's hottest spots.

Good luck!

 

[politicalslug]

Has anyone else found that the nTune autotune function in the performance tab doesn't work. For me all it does is attempt to increase the pciE bus, but it doesn't touch the fsb, memory, or gpu???

 

[lopri]

politicalslug, what CPU are you using? Are you using the stock HSF that comes with the CPU?

Forget about nTune for now. Current version is borked and it can't even provide correct readings.

 

[politicalslug]

Originally posted by: lopri
politicalslug, what CPU are you using? Are you using the stock HSF that comes with the CPU?

Forget about nTune for now. Current version is borked and it can't even provide correct readings.

Sorry, I forgot to mention that. I'm using an E6600 and a Zalman 9500. I also have the case door off and an 18inch floor fan pointing into the case. Is it normal for my 7900 gt (evga co superclock) to reach upwards of 89C. That sounds really high, like nearly boiling water high. Also, the memory is in linked mode, if that makes a difference. Memory's working well considering I only paid $150 for a 2048 package.

 

[3NF]

Originally posted by: politicalslug
I also have the case door off and an 18inch floor fan pointing into the case.

Is that the setup you are going to keep?? :Q

 

[skrewler2]

I got my 6600 stable at 3.8ghz last night, but the vcore was at 1.6 (set in BIOS) seems a little too high.. it got to 60C after 1 hr of Orthos, which seems too high.. I also needed to give it 1.6v to be stable at 1620fsb, which is odd. at lower voltages, the system can play games etc stably, but wont pass Orthos for more than 20 minutes..

I think 1.6v is a little too high

 

[Skott]

Lopri, 3.6 is a good overclock IMHO. Especially at those voltage settings. Be proud. Its a good OC.

Skrewler2, 3.8 is a very nice OC. 60C isnt bad. You are however at the point of not wanting to go higher temps if you can help it. 60C is the safe top end. I would say you need to see if you can get the voltage down some without losing that 3.8. It wont be easy with aircooling. 3.8 is pretty much top end on these mobos using aircooling from what I have read so far. What kind of hsf you running on it? Also did you pull the chipset heatsinks off and apply AS5? Thats a common mod many enthusiasts do to get lower temps.

 

[skrewler2]

Hmm.. today that O/C isn't stable after 45 minutes Orthos ... I don't want to increase voltage anymore.

I have the Tuniq Tower.

Switching my ram to Crucial Ballistix and running my board in sync mode seemed to help my overclocks a little bit, but nothing I want to keep b/c of the high voltage. I guess a watercooling kit or something would make it safe to run at those voltages? Hmm, speaking of watercooling, I really wonder what this board could do if the chipsets had waterblocks on them...

 

[politicalslug]

Originally posted by: 3NF

Originally posted by: politicalslug
I also have the case door off and an 18inch floor fan pointing into the case.

Is that the setup you are going to keep?? :Q

Yeah. I don't mind. It's relatively quite, and I'm relatively deaf. Besides, they line up well

 

[Skott]

Originally posted by: skrewler2
Hmm.. today that O/C isn't stable after 45 minutes Orthos ... I don't want to increase voltage anymore.

I have the Tuniq Tower.

Switching my ram to Crucial Ballistix and running my board in sync mode seemed to help my overclocks a little bit, but nothing I want to keep b/c of the high voltage. I guess a watercooling kit or something would make it safe to run at those voltages? Hmm, speaking of watercooling, I really wonder what this board could do if the chipsets had waterblocks on them...

If I'm not mistaken watercooling will get you to 4Ghz stable.