680i overclocking thread

[hifiking]

It looks just like the gum used on the 8800GTX core, it's more like foam. Had it been heavy and strong smelling I would give it a chance, but this stuff is too puffy.

[Khaotic]

Ok - SO
I'm looking to go with a 680i board.
If i go Asus, it will be NON Striker with P5N32-E SLi
I've tried reading over all the forums, but so far, it's going back and forth, adn I see a LOT of bad press. So, summarize - what are the pros and cons with the major leading 680i mobos. And i guess right now that comes down to the ASus P5N32-E and the eVGA.

[invicticus]

If the CPU vlts displayed in your pic is correct, you may want to reduce it. 1.6X v seems way to high for stock speeds. Try 1.35v and run test. If stable, reduce to 1.3v. This will help CPU run cooler too.

[Roger_D25]

Khaotic,
Back when I bought my eVGA 680i there wasn't as many other 680i boards available as there are now. With that said I would probably have gotten the Asus 680i (non-Striker) but not sure? It's a tough call since all these boards are faily new and havn't been sorted yet?

I see from your specs that your running an Asus P5B-Deluxe with an E6400 @ 3.6Ghz, how is that system working for you? I've heard many good things about that board, infact I was even thinking about scrapping the eVGA and trying that board out but I'm going to give the 680i some more time!

Eastcoasthandle,
Thanks for posting those pictures, much apprecaited! What 680i board are those pictures taken of, it looks to be one of the refernce versions? At the moment I've got my E6600 and my G80 watercooled but the N/S Bridges are still under stock cooling which isn't great especially once you start increaseing the voltage and frequencies!

I recenlty purchased DD's new 680i specific northbridge waterblock along with Evercools VC-RE southbridge cooler (which installs perfeclty using the eVGA's southbridge mounting holes)! I just hope that my Thermochil PA120.3 will be able to keep the CPU, GPU, and Northbridge at resonable temps?

I'm very curious if my N/S bridges will look similar to yours once I remove them? In any case I will take pictures of everything when I remove the stock coolers and install the new hardware. Plus I will post the before and after temps as accurately as I can using TAT, CoreTemp, and my DMM's non-contact temp probe. It will be difficult to accuratly monitor the N/S bridges temperatures because their not activly monitored in Windows, at least I don't know how to monitor these temps yet?

[GAR]

so far this is where i am at stable.

[SurlyPlatypus]

Hey Tyler,

It seems that I may be having the same problem with my E6600. I was oc'ing my CPU. Everything seemed stable at 3.7Ghz. I was able to run Orthos as well as 3DMark06 etc. I was playing 2142 for about 2 hours when my CPU heat alarms start sounding. when I exited 2142 the CPU showed 85C... in the ASUS PC Probe. I quickly shut the machine down and rebooted resetting my BIOS to all defaults. I ended up repairing XP due to what I believe is something totally unrelated (installed MBM just prior to the issue above, didn't reboot, so when I came back in windows would not boot.) I did a XP repair install and ever since windows only sees 1 CPU. Orthos, CPU-Z only show 1 core. no OC in BIOS now, although I can, same results. Bios says 2 cores, windows hardware in the control panel device manager shows 2 CPU's but my performance as well as CPU-Z and Orthos show only 1 CPU.

Did you ever get your issue corrected? Maybe it is the CPU itself that has a failed core?

hmmm ok this is really wierd....i just tried replaceing my other e6600 with the one that i was having troubles with...and it stopped having problems with the 1500 fsb strap.....very weird...i can go up to 1700 fsb with it now...however my cpu cant go that hight booting into windows....very very crazy....my first one is a 625a*** batch and the other that has the 1500 fsb strap problem is a 628b*** that makes me wonder if that could be what is having problems with this strap problem at 1500 fsb...anyone heard anything about this problem...and hey terminator have what batch is your cpu is a b by chance??? maybe those are that processors that have that problem?????


EDIT: sorry it really isnt a cpu strap its jsut a problem with the motherboard only using and alowing one core when it goes over 1500 fsb...a strap is a motherboard that cant go over a certain fsb

[sdumper]

Run->msconfig -> boot.ini -> advanced options -> numproc set to 2 and the restart...

[sdumper]

http://www.theinquirer.net/default.aspx?article=36448

Quote:

Nvidia issues Nforce 680i data loss fix

BIOS bodge


By Theo Valich: Monday 18 December 2006, 22:04

SOMETIMES, IT TAKES just a bit of a nudge to get things really moving. Sometimes a particular hardware component may come with a bit of a curse. Nvidia for example, didn't have much luck with now ever-present S-ATA standard. The first nForce4-SLI chipsets on the market came with corruption of data on hard drives. And history repeated itself.
The NForce 680i is without a doubt, better and more complete chipset than the 975X or P965. However, the whole "better" argument took a pasting when users started to report data corruption on SATA drives. And even worse, on RAID arrays, regardless of the type (0, 1, and 5).

It all came crashing down when Kylie penned an editorial at HardOCP.

Nvidia reacted quickly (just like they did in nF4-SLI time) and the fix has just been released.

Joe Darwin and Jacob Freeman from eVGA are working around the clock to explain everything to users which might need help.

The fix is nothing more than a beta revision of the BIOS with new RAID logic inside. If you have an EVGA 680i motherboard and experiencing data corruption on SATA drives, you can download the BIOS using our complementary L'INQ.

However, if you do not experience any problems, wait for the official BIOS rev which is coming in next couple of days. µ

[xgman]

The new bios didn't help me one bit. I officially give up on 680i. On to RD300 or 965 rev2.

[Blue078]

The new bios didn't help me one bit. I officially give up on 680i. On to RD300 or 965 rev2.

Did you format and do a fresh install after the flash?

[SmokeyTheBandit]

After weeks of struggling witch the EVGA reference and toying with a few sticks of ram i finally got the thing as stable as it for now will get

I'm getting more stability from the black memory banks than from the blue.
For some odd reason i can run very nice speeds up to DDR1000~1250 with my D9GMH and seem very solid/stable but it will freeze time after time ....

Today i got this result DDR850 1T @ 3-3-3-10 timings wich gives more bandwith then i got @ DDR1250

The only thing im going to do in the bios when i reboot again is saving my frikkin profile

http://img15.imgspot.com/u/06/352/17...1166567195.JPG

[elsupremo]

Hello guys,
My first post in this thread - I'm having some really annoying problems with this motherboard. I've spent way, I mean WAYYYY, too many hours trying to overclock this board to anything comparable to what I had before - I keep running into problems. This has been the hardest thing I've ever done, trying to get satisfied with this board. I used to have an Asus P5W DH, on which I was stable at 4.3 Ghz (orthos stable, benches, games, etc.). So far, the highest stable OC (hours of orthos, 32M superpi runs, etc.) I've been able to attain on this motherboard is 4.0 Ghz, and even then I had an odd freeze this morning that got me suspicious again.

My specs:
X6800 on Jinu117 phase change single stage (V2000).
EVGA 680i SLI, P23 Beta Bios
2 GB Gskill 800 HZ
2 8800GTX SLI
Xfi music
Boot drive: Raptor 74GB
PSU: PC Power Cooling 1 KW
Areca 4 port pcie RAID controller, with 4 300 GB Seagate drives in RAID 0
LG DVD burner

Like I said, I know my CPU can do 4.3 GHZ. However, if I select 4.2 or higher ghz on this EVGA, I run into weird problems. It seems to NOT matter what FSB I choose, I still run into the problem. For awhile last week, I thought I got stable at 4.2 ghz after I ran Orthos for several hours, and completed a couple of 1M Superpi runs. However, I was then playing Battlefield 1942 and the comp froze on me after 20-30 mins of play. Other, more graphically intense games (including 3dmark06) were no problem for me, but this old game froze the system. The only other program I've found that could easily recreate the freeze is Superpi, doing an 8M run or higher - something that takes a little while. The freezing is the problem I have at any of these higher overclocks. If I try 4.3 ghz, what I had before, it freezes much sooner, most of the time before windows can even load). When I use the term freeze, I mean it just stops dead, even during Windows load screen - the bar just stops. No restart, nothing. It just sits there.

I have tried dozens of voltage combos - some work better than others, but in the end the system still freezes. I have all spread spectrum settings disabled. I have thermal control and virtualization disabled, but disable bit enabled (as was mentioned earlier in the thread as a helpful setting).

The closest I've gotten to a stable system at 4.2 ghz (i'm even settling for 100 mhz less now than I had before) is with the following settings:
spread spectrum disabled
9 multiplier
1868 FSB
Memory - 800 mhz, auto timings (Playing it safe, if I ever get the system stable I'll mess with these)
Vcore - 1.7125 (This is overvolted. I only needed 1.7 for 4.3 ghz on the Asus, but I want to make sure CPU is not the problem here. Cpu heat is not the problem, as it stays well in the negative C's)
VFSB - 1.5
Vdimm - 2.1
Nforce spp - 1.55
nforce mcp - AUTO (seems to work best somehow, I've had NO LUCK with high MCP voltages)
HT spp-mcp - 1.55

With this setting, I was able to run 2 8M Superpi runs, but when I tried a 32M superpi run, the system froze after the fourth loop.

I upgraded last night to the P23 bios hoping it would help, that maybe I had the SATA bug and that was causing it, but the result I just mentioned was this morning and the freezing is still happening.

As I said before, the FSB doesn't seem to matter. I can run at 1780 FSB at 4 GHZ with mostly auto voltages, no problem. However, if I try 1720 FSB (with 10 multi) to get 4.2 Ghz, it will get into Windows, but will freeze during the superpi's, no matter what settings/voltages I try.

At this point, I am guessing there is some kind of heat issue in the chipset that is killing me when I overclock the CPU speeds. The issue doesn't seem to be related to the FSB, but just to the clock speed. In addition, it can pass Orthos with ease. Orthos doesn't elicit the problem, neither does 3dmark, netiher do most modern games. Only LARGE Superpi runs and a long time of playing Battlefield 1942. Does this make any sense to anybody? I highly respect how knowledgeable many people are on these forums, and I am sincerely hoping someone can help me. PLEASE, anybody with ideas, help me out. I am getting extremely frustrated by this board.

Thank you for reading my post, and I sincerely hope someone can help me.

[hifiking]

There is a great vdrop in the higher volts of the vcore. When i set 1.5, only get 1.46. And 1.775 gives 1.74. But you have tried them all i belive. Let us know if you find out anything, I have hit a wall too, but with 1.775 vcore I got to 4 ghz on my E6600.

[boogle]

Hi guys Sorry my first post is leeching info from your brains, rather than helping out - but you all are disturbingly skilled at PC hardware. I'm hoping you can shed some light on this, its been a long long time since I overclocked, so I'm probably missing something. I can't go over 395Mhz FSB, any higher and its just not stable. I've tried all sorts of voltages and timings, and it makes no difference.

Spec:
E6300 B2
Zalman 9500 LED
EVGA 680i P23 BETA BIOS
2GB Corsair Dominator C6400C4D
MSI 7900GTO (flashed to GTX)
OCZ 520W
3x Maxtor DiamondMax 10s 200GB (2x in RAID 0, one on its own)

Spread spectrums disabled. C1E, Speedstep, CPU thermal are all disabled.

I've set the main RAM timings manually, as well as the voltage; subtimings are all on auto (tried manual relaxed timings, made no diff). Currently I have all the voltages at stock (not auto), except for RAM at 2.1v, and CPU FSB at 1.3v. If I go to 1590 (397Mhz FSB), no matter what the voltage (any voltage, or combo of voltages) it's not stable.

I just don't understand it. Increasing SPP voltage (and/or CPU voltage) should increase the overclock surely? Even if its a small jump. But no, its like there's some sort of gremlin thats saying 'sorry - no more'. It doesn't fail Orthos, or any other stress test - it just freezes hard. Most times it'll reboot (no BSOD), other times it'll just freeze.

Any ideas? Thanks!

[ViperJohn]

Hi guys Sorry my first post is leeching info from your brains, rather than helping out - but you all are disturbingly skilled at PC hardware. I'm hoping you can shed some light on this, its been a long long time since I overclocked, so I'm probably missing something. I can't go over 395Mhz FSB, any higher and its just not stable. I've tried all sorts of voltages and timings, and it makes no difference.

Spec:
E6300 B2
Zalman 9500 LED
EVGA 680i P23 BETA BIOS
2GB Corsair Dominator C6400C4D
MSI 7900GTO (flashed to GTX)
OCZ 520W
3x Maxtor DiamondMax 10s 200GB (2x in RAID 0, one on its own)

Spread spectrums disabled. C1E, Speedstep, CPU thermal are all disabled.

I've set the main RAM timings manually, as well as the voltage; subtimings are all on auto (tried manual relaxed timings, made no diff). Currently I have all the voltages at stock (not auto), except for RAM at 2.1v, and CPU FSB at 1.3v. If I go to 1590 (397Mhz FSB), no matter what the voltage (any voltage, or combo of voltages) it's not stable.

I just don't understand it. Increasing SPP voltage (and/or CPU voltage) should increase the overclock surely? Even if its a small jump. But no, its like there's some sort of gremlin thats saying 'sorry - no more'. It doesn't fail Orthos, or any other stress test - it just freezes hard. Most times it'll reboot (no BSOD), other times it'll just freeze.

Any ideas? Thanks!

My EVGA has of whole areas/ranges of FSB settings that will cause the MB to boot into safe mode or to crash while loading/running Windows. The entire range from about 412Mhz (1648QDR) to about 425Mhz (1700QDR) would cause those problems. Above and below that range my MB work okays. There is another range of FSB settings lower down, that I know of, that causes the same issues for me.

I am at 429 x 9 now and working good and I can go up or down some on the FSB and have the MB still work. I did have to do the SPP volt mod and bumb the SPP volts up to 1.60 metered to run a 429fsb and change the NB/SB coolers to Viper custom units.

I also found the SPP runs (meters) higher than your bios setting. On my MB it was .048 higher than the bios SPP setting. I also found the 1.55 volt SPP setting in bios did not work. While 1.50 in bios gave 1.548 metered SPP volts the 1.55 setting in bios dropped the metered SPP volts clear down to 1.448..at least with the P21 bios. I did not check the P23 bios 1.55 volt SPP setting as I was already SPP volt modded at that time.

Makes it kind of hard to find your OC's when you have whole ranges of FSB settings where the MB won't work!!! You think you have hit a wall on the FSB that you can't go past but you can if you jump over (either up or down) the range(s) of FSB settings that give problems it will work.

Viper

[sdumper]

Is there a question in that post?

[ViperJohn]

Is there a question in that post?

Which post are you refering too??? If mine then no...just information on why you think
you might be hitting a FSB wall.

Viper

[boogle]

My EVGA has of whole areas/ranges of FSB settings that will cause the MB to boot into safe mode or to crash while loading/running Windows. The entire range from about 412Mhz (1648QDR) to about 425Mhz (1700QDR) would cause those problems. Above and below that range my MB work okays. There is another range of FSB settings lower down, that I know of, that causes the same issues for me.

I am at 429 x 9 now and working good and I can go up or down some on the FSB and have the MB still work. I did have to do the SPP volt mod and bumb the SPP volts up to 1.60 metered to run a 429fsb and change the NB/SB coolers to Viper custom units.

I also found the SPP runs (meters) higher than your bios setting. On my MB it was .048 higher than the bios SPP setting. I also found the 1.55 volt SPP setting in bios did not work. While 1.50 in bios gave 1.548 metered SPP volts the 1.55 setting in bios dropped the metered SPP volts clear down to 1.448..at least with the P21 bios. I did not check the P23 bios 1.55 volt SPP setting as I was already SPP volt modded at that time.

Makes it kind of hard to find your OC's when you have whole ranges of FSB settings where the MB won't work!!! You think you have hit a wall on the FSB that you can't go past but you can if you jump over (either up or down) the range(s) of FSB settings that give problems it will work.

Viper

Thanks for the info I tried 1716Mhz but it was no go Tried 1800, 1820, 1850, 1900, 2000 (no POST on this one). Both 6x and 7x multipliers. I'm really starting to gain an intense dislike of this board. It won't even fail Orthos, just freezes or reboots without BSOD. Gotta love the corruption that goes with that sort of sudden stop. Voltage doesn't matter whatsoever, other than too much SPP at certain speeds stops POST.

I tried knocking the RAM down to 667Mhz (unlinked). The result? Wouldn't even POST, and that was with manual timings set to what I have at 800Mhz. vdimm was 2.1v (again DDR2-800 speed). Its rediculous. Should have got a 965/975 mobo, then I could have just put in an FSB value, increased the MCH voltage and been well away.

[ViperJohn]

Thanks for the info I tried 1716Mhz but it was no go Tried 1800, 1820, 1850, 1900, 2000 (no POST on this one). Both 6x and 7x multipliers. I'm really starting to gain an intense dislike of this board. It won't even fail Orthos, just freezes or reboots without BSOD. Gotta love the corruption that goes with that sort of sudden stop. Voltage doesn't matter whatsoever, other than too much SPP at certain speeds stops POST.

I tried knocking the RAM down to 667Mhz (unlinked). The result? Wouldn't even POST, and that was with manual timings set to what I have at 800Mhz. vdimm was 2.1v (again DDR2-800 speed). Its rediculous. Should have got a 965/975 mobo, then I could have just put in an FSB value, increased the MCH voltage and been well away.

You may just have a MB that will not run. Even the ones that do are "twitchy" at best lol.

Your NB may simply need more SPP volts than you can get out of bios. I had to run 1.60 SPP (metered after Vmodding) to get to 429 (1716) on mine. Seems to be hugh amounts of FSB OC ability from one NB chip to another.

Viper

[elsupremo]

Update on my situation: Even though I didn't receive any direct responses in this thread, some kind soul sent me a Private Message suggesting I try some other FSB:RAM ratios. While the ratios didn't seem to solve the problem, playing around with the RAM DID help.

Specifically, I changed around some of the "advanced ram settings" in BIOS and I am able to run 32M Superpi runs pretty easily. Now, I have no idea what I'm doing with those settings. I'll tell you what I did first: I left the main timings at 4-4-4-12, with a 933 ram clock at 2.2V. Then I changed the advanced settings to 2-14-5-7, leaving the last setting Auto (7.8us). This was the change that allowed me to pass the 32M superpi runs. I've ran them repeatedly and I don't experience any more freezes! However, then I froze up while doing some odd things - When I opened Outlook and CPU-Z at the same time, it froze. Later, I was sending a message through ebay, and right after I clicked "Submit", it froze. I don't know what caused this, but I changed my settings again to see if it helps. My new settings for "advanced memory settings" are 2-11-5-6, leaving the 7.8us setting alone on auto.

I'm going to test these new settings for stability, but I'm surprised I had to change these memory settings around so much. Does this make sense to anybody? Why did this help me? I did not need this before on the Asus P5W.

Also, only at my new settings am I able to boot into Windows at 4.4ghz (something I could not do before, it would just freeze). However, I can't get superpi off, it just freezes again, so I think it's a similar situation again.

What should my advanced settings be for my 2 1GB Gskill 800 HZ sticks?

Any help is appreciated!

[Khaotic]

Khaotic,
Back when I bought my eVGA 680i there wasn't as many other 680i boards available as there are now. With that said I would probably have gotten the Asus 680i (non-Striker) but not sure? It's a tough call since all these boards are faily new and havn't been sorted yet?

I see from your specs that your running an Asus P5B-Deluxe with an E6400 @ 3.6Ghz, how is that system working for you? I've heard many good things about that board, infact I was even thinking about scrapping the eVGA and trying that board out but I'm going to give the 680i some more time!

Thanks for the reply - and sorry mine is so late.
My P5B WiFi Deluxe has been an AWESOME board. I understand all the "issues" with the vDroop, but the funny thing is, with the MANUAL overclock setting AND using AUTO as the vCPU setting - the voltage goes UP to the CPU as i increase the front side bus. Of course, CPU-Z stops reporting the correct voltage pretty quick, but the Asus probe shows that its at 1.55 +/- when I'm running with a 450-500FSB.
I've got an E6400 that will handle the high FSB just fine and I'm on Air cooling - Scythe INF series with a fairly quiet 120mm fan on the sink. I can get a 500Mhz front side out of the board/chip combo IF i lower the multi of the CPU. All in all, its been a REALLY fun board to play with.
I've got a eVGA 7950GX2 slapped on it, and I use it mainly for my LAN rig.

I'm looking to put a 680i into my MAIN rig to replace the P5W DH - becuase I want SLI without the mess of hacked drviers. AND i want a nice stable overclock to go with it. I have a Xeon 3060 on the way that has been sold to me by a fellow forumsman who says it will do 3.6 I expect that an Asus P5N32-E SLI or Asus striker Extrme should give up enough FSB to get a 6600/Xeon 3060 to 3.6 with little to no trouble while at the same time holding up an SLI setup. Is this a true statement everyone? I hope so - becuase that is the plan. I could REALLY R E A L L Y use some feedback on this one.

[eva2000]

My EVGA has of whole areas/ranges of FSB settings that will cause the MB to boot into safe mode or to crash while loading/running Windows. The entire range from about 412Mhz (1648QDR) to about 425Mhz (1700QDR) would cause those problems. Above and below that range my MB work okays. There is another range of FSB settings lower down, that I know of, that causes the same issues for me.

what happens in memtest86+ when you try those FSB ranges ? do you reboot or error out ? probably easier and safer to test in memtest first if the results would correspond with the windows crashes or safemode ?

[ViperJohn]

what happens in memtest86+ when you try those FSB ranges ? do you reboot or error out ? probably easier and safer to test in memtest first if the results would correspond with the windows crashes or safemode ?

Memtest will run until I shut it down as will P95.

Viper

[sdumper]

Maybe I missed it but has anyone posted a superpi yet at 4.5 or higher with the Striker or the EVGA board?

My striker was crippled with only one bank of DIMMs working so I never had a chance to push it for speed before I RMAed it. The replacement is due here in a week or so but I was hoping for some goals to shoot for.

[INFRNL]

can anyone here help me please? I have an evga 680i with an e6300 and I cannot get paft 1660 on the fsb. some settings wont even try to boot but others will boot to the windows screen and the scrolling thing just keeps scrolling. I have tried numerous settings and voltages with no luck. I think I may have a defective board.

I am also new to intel so I dont know what all the things are in the bios. can anyone display screenshots of their bios settings or tell me what their settings are?

Thanks
INFRNL