NVIDIA nForce 680i Solutions

[FrgMstr]

NVIDIA nForce 680i Solutions - NVIDIA answers some pointed questions about the recent nForce 680i chipset problems we have seen with their flagship motherboard chipset for Intel's Core 2 Duo processors. NVIDIA's own Drew Henry comes through with answers.

Hopefully NVIDIA has solved this 680i data corruption issue that a small but vocal percentage of 680i users were experiencing. If fixing it was as “easy” as a BIOS fix, (P23 BIOS) we have to say NVIDIA dodged a bullet although they will of course have to absorb some collateral damage. This misstep by NVIDIA surely taints the nForce 680i to some extent, but that will fade if the product stays on track and exhibits the quality traits we have witnessed in our testing. My personal experiences with the nForce 680i have been stellar to say the least. I think knowing that NVIDIA has been done their best to stay on top of this, working 24/7 until they had a fix, says something about the support behind their products.

 

[DoomRulz]

nV is getting a grip on the problem, that's excellent.

You said nForce reference boards are on the [H] Recommended List...where does a person purchase a ref board? Or did you mean the eVGA, BFG, etc... boards?

 

[Dan_D]

nV is getting a grip on the problem, that's excellent.

You said nForce reference boards are on the [H] Recommended List...where does a person purchase a ref board? Or did you mean the eVGA, BFG, etc... boards?

Yeah the eVGA and BFG boards are two examples of reference 680i boards.

 

[decapitator]

Yeah the eVGA and BFG boards are two examples of reference 680i boards.

What bull crap from Nvidia, I unloaded my evga 680i and ran for the hills. What about all the other problems like

Not compatible with PS2 mouse and keyboard,
Boot up will not load windows on random start ups.
Boot up without video and no error codes.
loss of ram in question due to over or undervolting of the board.
Fan noise after a week or so.
Intermittent network loss.
Run a successful OC only to wake up to a boot in safe mode as frequency needs reset.
BSOD's

And this is without raid

I went through three boards myself, one from Frys and two from newegg
I do not mind being a beta tester but not at these prices

Maybe nvidia will learn something from Abit and DFI on their up and coming 680i offerings or then again maybe they will be cursed with some off the same issues and then we can finally put the 680i chip set to rest.

Come on Intel SLI for the 775

 

[Arioch]

Well I am sure OCP will do it anyway but I went ahead and shot an email off to Aaron to make sure they use this bios or a newer version as they put my PC together.

My two concerns going in with my purchase were the sound problems with SLI that the P21 bios fixed and the SATA issues that this P23 bios addresses.

 

[an4rk1]

I was lucky enough not to have any problems, even with the initial bios and 4 sata drives raid'd. Very lucky it would seem. Although I couldn't get my X-Fi to install at all, but I think I just got a dodgy card so I went back to my trusty Audigy 4.

 

[magoo]

I agree with what the nVidia guy had to say about other boards.

I have the P5N32-E SLI and have had absolutely no problems. The board is simple to work with and has allowed me to OC without a problem.

Its too bad their own boards are causing people heartburn.

My advice, if you can find one,get the Asus board.

I am not bashing nVidia,trust me. I'm green through and through, but I have also found Asus to always be rock solid, too. I'm getting nice OCs on this board.....I just wish I knew more about memory and OCing it.

 

[Dan_D]

I've had zero problems out of my Striker Extreme. It's been everything a motherboard should be, aside from my initial overclocking issues. Which are now solved.

 

[magoo]

I've had zero problems out of my Striker Extreme. It's been everything a motherboard should be, aside from my initial overclocking issues. Which are now solved.

Have you gone past 400 FSB with the striker?????

 

[FrgMstr]

I've had zero problems out of my Striker Extreme. It's been everything a motherboard should be, aside from my initial overclocking issues. Which are now solved.

Well at the $400 price point, it stit did not show near the ocing that we hvae seen on the EVGA boards. Our second Striker did, so there seems to be no guarantees about that. I still can't suggest the board because of the cost.

 

[FrgMstr]

Have you gone past 400 FSB with the striker?????

Published last Friday.

 

[BinarySynapse]

What bull crap from Nvidia, I unloaded my evga 680i and ran for the hills. What about all the other problems like

Not compatible with PS2 mouse and keyboard,
Boot up will not load windows on random start ups.
Boot up without video and no error codes.
loss of ram in question due to over or undervolting of the board.
Fan noise after a week or so.
Intermittent network loss.
Run a successful OC only to wake up to a boot in safe mode as frequency needs reset.
BSOD's

And this is without raid

I went through three boards myself, one from Frys and two from newegg
I do not mind being a beta tester but not at these prices

Maybe nvidia will learn something from Abit and DFI on their up and coming 680i offerings or then again maybe they will be cursed with some off the same issues and then we can finally put the 680i chip set to rest.

Come on Intel SLI for the 775

All the issues you mentioned could be caused by the same timing problem that was mentioned in the interview. Except for the fan noise of course, You got a bum fan I guess..

 

[TheGameguru]

I've had zero problems out of my Striker Extreme. It's been everything a motherboard should be, aside from my initial overclocking issues. Which are now solved.

I thought you mentioned in another thread still having the sound crackling issue with certain games.

 

[Dan_D]

I thought you mentioned in another thread still having the sound crackling issue with certain games.

True. But only in a few games, I believe also that this is a video card drive issue, and NOT a motherboard issue. It only occurs when SLI is enabled.

 

[seehaw]

Way to pat yourselves ont he back.....maybe next time instead of kissing corporate ass so you don't loose your sponsors you could ask things like why are there BIOS holes all over the place on all your reference design motherboards? If you tout the overclocking potential of this board and every board seems to have holes with 4MB cache chips why didn't your QA team find this? Marketing a high end board to the high level gaming enthusiast would make one think that your QA department would test overclockability with 4MB cahce chips and realize there was a large disparity between results of 4MB and 2mB chips. Do you even actually have a QA department? Could your QA department possibly be just some monkeys with keyboards that get a banana for succesfullly hitting delete to enter the BIOS? Did you actually TEST the board at all before you shipped it? Do you consider it ok to hard launch what could be considered an Alpha product? Why can't you seem to develop an Intel chipset without dozens of wierd compatibility issues? I could go on for days.......

Seriously I'm not impressed [H] you used to have more credibility. And I know the next line coming is we talked to NVidia bla bla bla we were the only ones retracting our stories yadda yadda. The same dog and pony show you've used for years is getting old. You really had the chance to ask some questions everyone is curious about but didn't. I guess even though you try to act like it's all about us it really is business at the end of the day. I can understand that being a businessman myself just be honest with us and yourselves and I think it'll be easier to swallow.

 

[BinarySynapse]

Maybe they didn't ask what you want because the data corruption issue is far more important than how well you can void your warranty.

Way to pat yourselves ont he back.....maybe next time instead of kissing corporate ass so you don't loose your sponsors you could ask things like why are there BIOS holes all over the place on all your reference design motherboards? If you tout the overclocking potential of this board and every board seems to have holes with 4MB cache chips why didn't your QA team find this? Marketing a high end board to the high level gaming enthusiast would make one think that your QA department would test overclockability with 4MB cahce chips and realize there was a large disparity between results of 4MB and 2mB chips. Do you ever actually have a QA department? Could your QA department possibly be just some monkeys with keyboards that get a banana for succesfullly hitting delete to enter the BIOS? Did you actually TEST the board at all before you shipped it? Do you consider it ok to hard launch what could be considered an Alpha product. Why can't you seem to develop an Intel chipset without dozens of wierd compatibility issues? I could go on for days.......

Seriously I'm not impressed [H] you used to have more credibility. And I know the next line coming is we talked to NVidia bla bla bla we were the only ones retracting our stories yadda yadda. The same dog and pony show you've used for yours is getting old. You really had the chance to ask some questions everyone is curious about but didn't. I guess even though you try to act like it's all about us it really is business at the end of the day. I can understand that being a businessman myself just be honest with us and yourselves and I think it'll be easier to swallow.

 

[coachjohn]

recommendation on the 680i boards, you wrote:

"Shooting from the hip on this, and I am no driver engineer, but this sounds like a hardware problem to me. I hate to say it, but there seems to be a chipset problem here that is not apparent in every piece of silicon. I hope I am wrong and it is just a driver issue that can be fixed through a new driver drop, but I am not hopeful of this. After seeing Dan take his 680i build that was causing him severe problems, change out motherboards, and his problems then going away, it is hard for me to think that a driver caused the issue."

Does the explanation that "The problem that some users were experiencing related to signal timings on the motherboard" seem consistent with your original thoughts...could signal timings cause all of the types of problems reported? There have been a few reports of an option in the BIOS (Spread Spectrum) being disabled by the new BIOS - is this true or peculiar to some systems and configurations? If someone hasn't had any problems to date...is the logical recommendation that they flash the board with the final release of the new BIOS when it becomes available?

I've been fortunate...no major problems to date...but I've been following these threads with interest, and the comment from the interview that "there are statistical variations in electrical characteristics between boards, which in some cases, can mask the underlying issue" caught my attention as a possible red flag...the implication is that the problem is there on all systems but just hasn't revealed itself yet.

 

[seehaw]

Maybe they didn't ask what you want because the data corruption issue is far more important than how well you can void your warranty.

Than a board selling at a higher price points specifically because it should be a great overclocker and even says that on their product brief when in fact it's barely functioning. Oh maybe the fact the papers also say it's a breeze to overclock and they encouraged that be looked at should be ignored too because you can only ask about one thing at a time in your world. So when a reporter interviews someone they should never cover all the topics surrounding a story they should pick one and stick with it. I guess i see your point. Well let me just apologize, [H] I apologize for ever expecting you or anyone else to thoroughly cover all the issues with a product during an interview with the product designer and maker. That was awfully contmeptous of me. I can't believe I ever expected reporters/reviewers to fully examine a product they recommend and ask why stuff is broken with it. God I'm out of line. Hope you feel better ryan.

 

[BinarySynapse]

Than a board selling at a higher price points specifically because it should be a great overclocker and even says that on their product brief when in fact it's barely functioning. Oh maybe the fact the papers also say it's a breeze to overclock and they encouraged that be looked at should be ignored too because you can only ask about one thing at a time in your world. So when a reporter interviews someone they should never cover all the topics surrounding a story they should pick one and stick with it. I guess i see your point. Well let me just apologize, [H] I apologize for ever expecting you or anyone else to thoroughly cover all the issues with a product during an interview with the product designer and maker. That was awfully contmeptous of me. I can't believe I ever expected reporters/reviewers to fully examine a product they recommend and ask why stuff is broken with it. God I'm out of line. Hope you feel better ryan.

Hmm, why don't you call up Nvidia and ask them yourself and report back to us on you findings?


Oh and by the way, holes are present on just about every chipset that can get to 400+ FSB's. This include the 975X, P965, and even the 680i. Not hitting a 375FSB when 401 works just fine isn't a show stopper when you XP installation gets corrupted every other time you boot at stock speeds. And who knows, maybe the holes are caused by the very same timing problems that is corrupting data.

 

[magoo]

I would have liked to have seen more questions and answers, too. Maybe address other "bugs" that nVidia is hearing about.
But it seems like this one issue has been very bad and obviously if your system is useless all the cool OC options in the world are just plain useless.

For one, I read the nVidia OC guide before I bought my board, I followed their suggestions to the letter and man does my board perform. I think alot of people just need to read and research before they condemn.

 

[decapitator]

I just love how people with working boards (for now) act with the ones who don't have working boards This hole nvidia roll out campaign got me caught up in it as well, We have the nvidia Mother Board, the nvidia SLI ram, the nvidia Zalman matching fan and nvidia 8800 series cards all released on the same day with the exception of the fan and ram to the fan fare of an opening movie

 

[decapitator]

Read and research all you want, it wont fix a broken board that is not ready for prime time.

 

[Hurin]

Read and research all you want, it wont fix a broken board that is not ready for prime time.

Looks like someone is on a crusade.

Aside from the occasional beep due to my PS/2 keyboard. . . I'm having zero issues with the high-end rig in my sig.

So, chalk yourself up to the ever-decreasing number of people having show-stopping problems with this board. But, that number is decreasing rapidly. Your personal experience is not necessarily definitive. But, judging by how vocal you suddenly are, it appears that you want it to be.

 

[grdh20]

This bios didn't help me one bit with sata corruption. After 4 680i boards I'm done with 680i.

 

[PeteWalker74]

I believe NV has responded extremely well to this, fixing (hopefully) the issue in a very short amount of time. I don't think anyone should be bashing them at all, or commenting that this tarnishes the image of the board. I didn't have this issue with my eVGA board. Even if I did... sure I'd be pissed, but it was fixed in such a short amount of time that it would have barely effected me.

 

[FreedomHDM]

I'm fairly happy with my 680i SLi board from evga. I got this thing from newegg about two weeks ago and it easily overclocked with the e6600. However, about a week after I overclocked I started getting some problems. My computer would freeze up and become generally unstable while running games. It turned out to be a voltage problem. I thought I tweaked everything out perfectly, but I guess the CPU needed more juice. It's strange because it worked fine for about a week.

Anyways I reset all the BIOS defaults and re-overclocked. It's been 3 days so far, and everything seems fine again. Hopefully it won't start giving me problems again. Is this a BIOS version issue? I only have the version it came with on CD.


Edit: I just discovered that I had the PS2 Laggy/jumpy mouse problem. Yay. Luckily I have a USB mouse. But, unfortunately I do not have a floppy drive or usb key so I can not install any bios updates. Idiotic.

 

[TheGameguru]

I believe NV has responded extremely well to this, fixing (hopefully) the issue in a very short amount of time. I don't think anyone should be bashing them at all, or commenting that this tarnishes the image of the board. I didn't have this issue with my eVGA board. Even if I did... sure I'd be pissed, but it was fixed in such a short amount of time that it would have barely effected me.

That makes zero sense.. whatever/wherever the issue or fault may lay in the end.. clearly this product was rushed to market. It's bad enough that we allowed buggy software to become the norm but now buggy (high end at that) PC gear is simply going to be glossed over?

I sure hope not.

 

[Hurin]

This bios didn't help me one bit with sata corruption. After 4 680i boards I'm done with 680i.

Did you flash via Floppy?

Did you clear the CMOS via jumper after flashing?

Did you format the hard drive and re-install Windows?

All three of those should be done before you discount this as a good fix.

 

[magoo]

Read and research all you want, it wont fix a broken board that is not ready for prime time.

I dont attempt to rub anything in anybody's face........I have a board that works just fine, thank you, but I did my homework and didnt buy one on day one of the release either.

Do you or did you have a 680i board or are you just trying to start something????

I'm sorry people are having problems and it seems to me that nVidia is doing all they can to fix the problem.

I introduce my personal experience to let people know that there are good 680i boards out there, and what mine is capable of, if they are considering buying one.

You seem a bit bitter.........

 

[Baer]

I , for one, am pleased that NV seems to be on top of things. This chipset has so much promise and may be breaking enough ground that some glitches are expected. NV's admission and effort are reassuring. I see that New Egg has now put a new ETA date on their site for availability of the Striker 12/25. Perhaps a Christmas present for those of us that have been trying to get one.
While I am sure that there will be other issues show up I am now more confident that NVIDIA is aware of the issues and even if they have not fixed them all as yet, they are working on it.

 

[Kuja]

FWIW, I recently got a DOA 680i. Talk about dead - no beeps, not even any diagnostic codes. The eVGA guy I got a hold of on the phone seemed not to have heard about too many boards like that, but the internet disagrees. He also told me to RMA through newegg, because it'd be faster than through eVGA.

I'm pretty sure this isn't really eVGA's fault, and that they're taking a huge bullet for nVidia. I mean, it's a reference board, right? If the second 680i that I get is also FUBAR'd, there'll be hell to pay. BTW - nothing but good things to say about the rest of the evga lineup, I've owned an NF41, NF43 (Socket 939 motherboards) and several of the geforce offerings.

 

[FrgMstr]

Way to pat yourselves ont he back.....maybe next time instead of kissing corporate ass so you don't loose your sponsors you could ask things like why are there BIOS holes all over the place on all your reference design motherboards? If you tout the overclocking potential of this board and every board seems to have holes with 4MB cache chips why didn't your QA team find this? Marketing a high end board to the high level gaming enthusiast would make one think that your QA department would test overclockability with 4MB cahce chips and realize there was a large disparity between results of 4MB and 2mB chips. Do you even actually have a QA department? Could your QA department possibly be just some monkeys with keyboards that get a banana for succesfullly hitting delete to enter the BIOS? Did you actually TEST the board at all before you shipped it? Do you consider it ok to hard launch what could be considered an Alpha product? Why can't you seem to develop an Intel chipset without dozens of wierd compatibility issues? I could go on for days.......

Seriously I'm not impressed [H] you used to have more credibility. And I know the next line coming is we talked to NVidia bla bla bla we were the only ones retracting our stories yadda yadda. The same dog and pony show you've used for years is getting old. You really had the chance to ask some questions everyone is curious about but didn't. I guess even though you try to act like it's all about us it really is business at the end of the day. I can understand that being a businessman myself just be honest with us and yourselves and I think it'll be easier to swallow.

Then I woud highly sugget you go elsewhere for your hardware research needs. I am sorry we do not live up to your standards. Your thoughts are noted. Please do not carry this thread off topic. Should you wish to discuss my business practices, please take it up via email.

 

[RAutrey]

To get back on topic: I think this is nothing more than a component issue. It's not even design related. ASUS boards are working. NV builds were not. Same design. Different manufacturing facilities. NV probably relaxed the timings a bit to settle out a component critical to clock generation. I would like to see if there is any measurable difference in performance between an ASUS board with a prefixed BIOS and a NV ref board with the fix.

 

[FrgMstr]

FWIW, I recently got a DOA 680i. Talk about dead - no beeps, not even any diagnostic codes. The eVGA guy I got a hold of on the phone seemed not to have heard about too many boards like that, but the internet disagrees. He also told me to RMA through newegg, because it'd be faster than through eVGA.

I'm pretty sure this isn't really eVGA's fault, and that they're taking a huge bullet for nVidia. I mean, it's a reference board, right? If the second 680i that I get is also FUBAR'd, there'll be hell to pay. BTW - nothing but good things to say about the rest of the evga lineup, I've owned an NF41, NF43 (Socket 939 motherboards) and several of the geforce offerings.

I wish I had a dollar for every dead motherboard I have seen here. Surely it is frustrating, but says little as to the overall quality of the ful series of motherboards.

EVGA fought pretty hard to get the etailer exclusive on this, guess you need to be careful what you wish for.

 

[FrgMstr]

Thank you Kyle for keeping me from getting banned. I was about to flame the boy with a 30 megaton weapon. I have been thinking about getting the EVGA board for a while and am going to go pick one up tonight. When brought [H]ard's integrity into question, it really made me mad because I remember the year that your integrity caused you to have to move the [H]ard OCP workshop to the Adamsmark because of pressures being put on you to sugar coat some things with a certain manufacturer and you refused to do it. BTW, if I get a bad board I'll email you to let you flash it ti see if that will fix it. You still around the Dallas area? (PM me)

Thank you sir! And yes, I am in Plano and can get you my UPS number or swing by and get it. I would happily take a bad one off your hands.

 

[FrgMstr]

To get back on topic: I think this is nothing more than a component issue. It's not even design related. ASUS boards are working. NV builds were not. Same design. Different manufacturing facilities. NV probably relaxed the timings a bit to settle out a component critical to clock generation. I would like to see if there is any measurable difference in performance between an ASUS board with a prefixed BIOS and a NV ref board with the fix.

The Asus board is NOT of the same design as the reference board.

 

[Stoney]

Well the fact that they gave an interview and then "came clean" and admitted there was a problem, is a step in the right direction.
For the users that are having problems, i feel for yea.
I got the Evga i680 mobo and it's still in shrink wrap, so haven't had a chance to put my system together. (real life issues)
But it's nice to know when i get around to it there will be support behind my new build.
And in the event the mobo goes south the thing has a lifetime warranty, so "No worries"

As far as [H]ard0 Ocp's reporting abilities go...nothing has changed my mind that there anything else but "Top notch- stand up dudes".
1. Their review of the i680 was reported as found.
2. when problems started showing up, they revisted the review and reported that also.
3. They got Nvidia to come in for an interview and face the music.
(haven't seen that before, good job)

So Kyle & Dan, kudos on your efforts.

To all the angry people out there, man go find a girlfriend-"Yea-all need some lovin"

 

[Baer]

Kyle,
You did good and handled that well.
This is a very valuable resorce for those of us that are willing to take the risks and deal with the bleeding edge issues. I only wish I had found this site years ago.
Excellent job, Excellent!
When I saw your scoop on the home page this morning I was elated.

I have been in leading edge electronics since vacuum tubes and latching relays. We used to have to wait for a monthly magazine to get an update on a new product. [H] is a very impressive site. Great job.

Now back on topic.... sorry

 

[Diseaseboy]

Now if we can only get NVIDIA to give us an answer to the poor 8800 drivers too.........

 

[Dan_D]

To get back on topic: I think this is nothing more than a component issue. It's not even design related. ASUS boards are working. NV builds were not. Same design. Different manufacturing facilities. NV probably relaxed the timings a bit to settle out a component critical to clock generation. I would like to see if there is any measurable difference in performance between an ASUS board with a prefixed BIOS and a NV ref board with the fix.

The design of the two boards, and the choices of components are completely different. They may look similar on the surface, differing in cooling method and color choices, but they are totally and completely different. The ASUS board has all it's onboard audio on a daughter card that uses a proprietary PCIe type interface, the ASUS board also has SATA ports, front panel connectors and everything else in different locations and the capacitors and other components mounted on the board are made by different manufacturers than the eVGA board. In no form or fashion are the boards the same. Aside from their basic layout, (from a glance) and their chipsets, they couldn't be more different.

There is no appreciable difference in performance between the two boards. There is a 1-3% variance, but that's what it is, a variance. The difference is within an acceptable margin of error for the tests.

Try benchmarking your board. It won't score the exact same numbers every time. It should vary 1-3%. If it shows more variation than that, something is wrong with your system.