Well, I killed another 680i SLI board.

[Dan_D]

Well, I just killed yet another EVGA 680i SLI board.

I was having increasing stability issues, returned the board to stock, that didn't help, so I flashed it with the latest BIOS and now, it's BSOD hell. Swapped the ram, now the thing won't post anymore.

So I'm installing the EVGA Black Pearl as I type this.

 

[Arcygenical]

Yay... I really hate the 680i boards. Some work fine, but when they fail, they fail magnificently.

 

[Uberbob102000]

Oh yea... My gigabyte just died spectacularly like Dan's about 3 days ago.

 

[Dan_D]

Yay... I really hate the 680i boards. Some work fine, but when they fail, they fail magnificently.

Well I've had 11 EVGA boards, and one ASUS board. The ASUS board still works, but it's a pitiful overclocker, and it hates most of the RAM I've ever tried in it.

I went through three of them when they first came out. I had experienced the SATA corruption and audio issues. I gave up on them for awhile and used the ASUS. Then I gave the EVGA another when my ASUS refused to accept the new higher performance RAM I purchased. Then I retired that EVGA board to another computer when I needed quad core overclockability. After that is when it got real ugly. I had two DOA boards and I physically broke one. That was my fault, but the DOA's obviously weren't. I fried another one in a coolant spill, (also my fault) and then I got another one that quit working in dual channel mode, and wouldn't post with more than one DIMM slot populated. Then I installed number 10 here, and well, it went down as I described above. So now I'm installing the Black Pearl, which I kept from a previous review I did.

I really hope that the 780i SLI chipset based boards are better than these. Replacing these things, and taking my machine apart is getting old. To EVGA's credit, I don't think this is primarily their fault, I think it's more of a NVIDIA design issue than anything, or at least to some degree, chipset related. It seems that all the 680i SLI boards have major issues compared to other high end solutions. I'd go with a 680i LT if it had a third PCIe slot. I need it for my SATA/SAS RAID controller.

 

[spaceman]

Yeah. I went with the MSI 650 SLI Plat 8 months ago and have had zero issues so far. Weird. What is it with the 680s that makes them so failure prone?

 

[Dan_D]

Yeah. I went with the MSI 650 SLI Plat 8 months ago and have had zero issues so far. Weird. What is it with the 680s that makes them so failure prone?

I have no idea. On most of the boards I've had, I've noticed that just prior to death, the north bridge temps get way out of whack (increasing by 20c) and then BAM.............toast.

 

[BillParrish]

My understanding (and I admit I don't keep up much with Nvidia chipsets) is that the 780i is basically a 680i with some add on chip to supply the PCI-2 2.0 ? Actually I just took another look see and some sites are claiming it IS just a 680i with the extra chip.

Soooooo IF (caps intentional) this is true (takes grain of salt), based on the past record of issues and ram toasting, I would be surprised if the 780i is a lot better. One would hope that some of the issues have been resolved but since Nvidia does not publicly publish the details of there chipset specs, I would want some very specific answers to questions as to what exactly has changed with the design aside from adding the extra PCi-e controller. Measurements of the voltage at the dram slots vs what is set in the bios would be entertaining as well.

I look forward to the review.

And I am very sorry for your loss, bummer.

 

[LOCO LAPTOP]

I gave up on the 680I. I really did. I got sick of RMA's every 2 weeks. I even had the NB on water and still died. Im not gonna be a new NV chipset unless the new one is better. As of last week I had X38 love.

 

[Q-BZ]

Well I've had 11 EVGA boards, and one ASUS board. The ASUS board still works, but it's a pitiful overclocker, and it hates most of the RAM I've ever tried in it.

I know I don't want to bother with these 680i boards and probably not the 780i boards, either.

Which Asus board were you referring to here?

 

[northrop]

my EVGA 680i is 11 months old now, and it never gave me any trouble at all! guess i'm a lucky sob.

 

[Dan_D]

My understanding (and I admit I don't keep up much with Nvidia chipsets) is that the 780i is basically a 680i with some add on chip to supply the PCI-2 2.0 ? Actually I just took another look see and some sites are claiming it IS just a 680i with the extra chip.

This is my understanding as well, but I can't say for sure as I have yet to see a 780i SLI board in person, nor have I had the chance to work with a sample yet.

Soooooo IF (caps intentional) this is true (takes grain of salt), based on the past record of issues and ram toasting, I would be surprised if the 780i is a lot better. One would hope that some of the issues have been resolved but since Nvidia does not publicly publish the details of there chipset specs, I would want some very specific answers to questions as to what exactly has changed with the design aside from adding the extra PCi-e controller. Measurements of the voltage at the dram slots vs what is set in the bios would be entertaining as well.

Well if and when I get ahold of a 780i SLI board, I will be sure to scrutinize it very carefully. Especially in regard to voltage readings. That said I wouldn't be surprised at all if the 780i SLI were actually a much improved design. All the boards that NVIDIA designed after the 680i SLI boards were all far better in terms of reliability and stability. None of those boards had the teething issues the 680i SLI boards did. So I'd be shocked if the 780i SLI board weren't simply amazing. I guess I'll find out when they are closer to a release date for them, or just shortly after they are released.

I look forward to the review.

And I am very sorry for your loss, bummer.

Not a big deal really, I've lost a bunch of these boards, but all of them except for the one that just died were returned to the retailer and exchanged. So I'm not out any extra cash or anything. I think I'm just pushing them too hard with the quad core or something and that's why they keep dying. I've got one board that's nearly a year old running with a dual core CPU and it's overclocked, and running strong. That's the system I'm actually typing on right now, and it's solid. My girlfriend plays games on it all the time without issue.

I gave up on the 680I. I really did. I got sick of RMA's every 2 weeks. I even had the NB on water and still died. Im not gonna be a new NV chipset unless the new one is better. As of last week I had X38 love.

I'd love to go X38, but I don't think I can live without SLI support.

I know I don't want to bother with these 680i boards and probably not the 780i boards, either.

Which Asus board were you referring to here?

The ASUS Striker Extreme. Also I've got a friend who has a P5N32-E SLI and it's been unreliable too.

my EVGA 680i is 11 months old now, and it never gave me any trouble at all! guess i'm a lucky sob.

Well as I said, I've got one 680i SLI board and it's problem free. One I broke a capacitor off of during a botched water block install was sold to a friend after testing, and it's working awesome to this day. So not all of them are bad, but the majority of the ones I've had were pretty bad. Though usually they all died in less than 30 days if they weren't DOA. This one is two or three months old.

 

[Nick_Leo]

my 680i is alright but i do get these random ass bsod and they piss me off so much, other than that and the super hot nb its treated me well.

 

[Dan_D]

I know I don't want to bother with these 680i boards and probably not the 780i boards, either.

Which Asus board were you referring to here?

my EVGA 680i is 11 months old now, and it never gave me any trouble at all! guess i'm a lucky sob.

my 680i is alright but i do get these random ass bsod and they piss me off so much, other than that and the super hot nb its treated me well.

That's been my experience as well. Mines been pretty solid until about two days ago. It started BSOD'ing on me and hard-locking while playing games. Today I couldn't play COD4 for more than a few minutes before the program would lockup, and the system BSOD. I flashed the BIOS hoping to resolve the problems, but that only made things worse. I hadn't been able to stabalize it at all, not even enough to get into the OS, and then bam, I swapped memory, and it wouldn't boot after that. Not even the one stick of ram in the far slot trick would ressurrect this thing. So I gave up and finally I'm installing the Black Peal Edition board I've had sitting around for 2 months now.

I'm also going to create two loops in my PC this time. One for the video cards, and one for the board and CPU. For the short term I'll be using the Nautilus 500 for the CPU and motherboard blocks, and the Swiftech setup will handle the video cards alone. This should hopefully help to further control north bridge temperatures. Plus it should make things quieter as the SB fan I had on my other board was quite loud. The other reason for going dual loop here is so that I can do motherboard replacements, or video card replacements having to only drain one loop instead of the whole thing.

 

[GreenGoose]

After 4 sets of burned up RAM and two motherboards, I gave up on 680i. Intel is the only way to go for me.

 

[BUFF]

I've read too that 780i SLI is just 680i SLI with a PCI-E2.0 chip but eVGA say that their 680i SLI won't support Yorkfields but that the 780i will.
So, either there is more to 680i/780i changes or it's a marketing decision to force upgrades ...

 

[BrainEater]

I feel your pain Dan.

Hope mine keeps running for a while.

 

[Dan_D]

Well just a little update, it seems more was going on than a dead motherboard. In fact, it doesn't look like that motherboard died afterall. I replaced the board, and I had the same BSOD'ing issues. To make a long story short, I ended up replacing the RAM and now the system stabalized. Here is the worst part, somehow my RAID controller seems to have really screwed up during all of this. I ended up reinstalling Windows and I couldn't see the array at all. Everything seems find according to the card BIOS, but Windows can not and won't see the array. I threw in a spare hard drive to get going again and the system is working flawlessly, but the MegaRAID shows up with an exlamation point in device manager. None of the drivers from LSI seem to be working.

Basically I only figured all of this out around 2AM. So I essentially spend over 12 hours working on this thing. Granted much of that was messing with the water cooling. In any case the box is working great, the new Black Peal and split loops are giving me much better cooling results, which is fantastic. I've still got to solve my SAS/RAID card problem and get my Raptors going again. I'd really hate to go back to the NVRAID controller. I'm sure many of you can understand.

As for the old board, it seems to be working, but it's still having it's USB problems and other issues I have always had with it, so I needed to swap it anyway, but it looks like RAM and a corrupted OS (probably due to the RAM's instability) were the real cause of my issue. So while I thought I had escaped the whole 680i SLI chipset RAM frying business, it would seem that I didn't.

 

[stain.]

my 680i has been running fine. I've had a TON of ram issues (corsair 8500c5d), and have returned the ram a lot. I got tired of dealing with it, and have been on OCZ ram for about a week and a half now, love them but I'm waiting to get my 4x1gb mushkin in.

the board's been solid for me.

 

[GreenGoose]

my 680i has been running fine. I've had a TON of ram issues (corsair 8500c5d), and have returned the ram a lot. I got tired of dealing with it, and have been on OCZ ram for about a week and a half now, love them but I'm waiting to get my 4x1gb mushkin in.

the board's been solid for me.

You are having a 'TON' of ram issues because your 680i is overvolting the sticks and killing them. It is a well documented issue. It's just a matter of time before it burns your OCZ up too. You call that a solid board?

 

[Ihaveworms]

You are having a 'TON' of ram issues because your 680i is overvolting the sticks and killing them. It is a well documented issue. It's just a matter of time before it burns your OCZ up too. You call that a solid board?

Yeah, it seems to love overvolting RAM. Thats why I have mine set on 1.85v to the RAM.

 

[nst6563]

I just recently got a 680i board with a q6600 and am now just reading about all these problems. Making me nervous to say the least. I'll google for this as well, but is there a point on the motherboard that I can read the voltage going to the memory? I can look in the bios, but so far it always shows the correct voltage. This overvolting issue that is mentioned... is it short spikes in voltage or a constant overvolt?

thx

 

[GreenGoose]

I just recently got a 680i board with a q6600 and am now just reading about all these problems. Making me nervous to say the least. I'll google for this as well, but is there a point on the motherboard that I can read the voltage going to the memory? I can look in the bios, but so far it always shows the correct voltage. This overvolting issue that is mentioned... is it short spikes in voltage or a constant overvolt?

thx

Do not believe what voltage the motherboard is reporting. I burned up 4 sets up Corsair even after intentionally setting the voltage lower. It will eventually burn them out regardless of what you set it to.

 

[BrainEater]

I just recently got a 680i board with a q6600 and am now just reading about all these problems. Making me nervous to say the least. I'll google for this as well, but is there a point on the motherboard that I can read the voltage going to the memory? I can look in the bios, but so far it always shows the correct voltage. This overvolting issue that is mentioned... is it short spikes in voltage or a constant overvolt?

thx

here : dead 680i

125,000 hits....this doesnt include the '680i eats ram' ones either.

srry mang.

 

[nst6563]

Do not believe what voltage the motherboard is reporting. I burned up 4 sets up Corsair even after intentionally setting the voltage lower. It will eventually burn them out regardless of what you set it to.

I know, which is why i asked if there was a spot that it could be measured from on the board.

guess I'll have to do some googling... good thing I'm still within my 30 day return period.

 

[Dan_D]

You are having a 'TON' of ram issues because your 680i is overvolting the sticks and killing them. It is a well documented issue. It's just a matter of time before it burns your OCZ up too. You call that a solid board?

Not necessarily true. I've got two sets of OCZ Flex XLC PC2-9200 (DDR2-1150MHz) that need to be run at 2.3v on the EVGA 680i SLI boards and one set has been running for more than six months at that voltage without any problems. I've got a second set that's working quite well too at the same settings and also has been doing so for some time. OCZ even warranties this RAM being run at 2.3v.

my 680i has been running fine. I've had a TON of ram issues (corsair 8500c5d), and have returned the ram a lot. I got tired of dealing with it, and have been on OCZ ram for about a week and a half now, love them but I'm waiting to get my 4x1gb mushkin in.

the board's been solid for me.

I've had a lot of compatibility problems with Mushkin RAM over the years. Beyond that good luck with running a 4x1GB configuration on any 680i SLI board. It's a bitch to get working and flat out can't be done with some modules.

Yeah, it seems to love overvolting RAM. Thats why I have mine set on 1.85v to the RAM.

Not all ram can be run at 1.85v. My Patriot modules need 2.0v and my OCZ modules need to be run at 2.1v-2.3v depending on the clock speed they are set at.

I just recently got a 680i board with a q6600 and am now just reading about all these problems. Making me nervous to say the least. I'll google for this as well, but is there a point on the motherboard that I can read the voltage going to the memory? I can look in the bios, but so far it always shows the correct voltage. This overvolting issue that is mentioned... is it short spikes in voltage or a constant overvolt?

thx

The failures seem to be more common with quad cores on the 680i SLI boards than anything. Running a dual core processors yields better results on these boards. Unfortunately.

Do not believe what voltage the motherboard is reporting. I burned up 4 sets up Corsair even after intentionally setting the voltage lower. It will eventually burn them out regardless of what you set it to.

I don't believe this. While there are certainly problems burning up memory modules on the 680i boards, this isn't necessarily inevitable.

 

[magoo]

Maybe I should knock on wood or something.
I have 4x1 GB of Corsair XMS 6400 C4 on my P5N32-SLI board and have not had any problems for 1 solid year.

I have my RAM unlinked and let the board select the voltage.

 

[BillParrish]

I know, which is why i asked if there was a spot that it could be measured from on the board.

guess I'll have to do some googling... good thing I'm still within my 30 day return period.

http://pinouts.ru/Memory/dimm_ddr2_pinout.shtml

Looks like a couple of voltages involved. I do not see why one could not measure them all on an empty socket for starters and then investigate exactly what each voltage is susposed to be in relation to the "main" voltage which is VDD pin 64 and 184. the pins next to the key on the "short" side.

Dammit I hate post like yours, now I will probally spend all night looking at memory data sheets and JEDEC standards instead of playing my game.

This seems to cover most of it or at least give a start to see if default voltages are "in the ball park"
http://download.micron.com/pdf/datasheets/dram/ddr2/512MbDDR2.pdf


Table 20: Recommended DC Operating Conditions (SSTL_18)
All voltages referenced to VSS
Parameter Symbol Min Nom Max Units Notes
Supply voltage VDD 1.7 1.8 1.9 V 1, 5
VDDL supply voltage VDDL 1.7 1.8 1.9 V 4, 5
I/O supply voltage VDDQ 1.7 1.8 1.9 V 4, 5
I/O reference voltage VREF(DC) 0.49 x VDDQ 0.50 x VDDQ 0.51 X VDDQ V 2
I/O termination voltage (system) VTT VREF(DC) - 40 VREF(DC) VREF(DC) + 40 mV 3

Notes: 1. VDD and VDDQ must track each other. VDDQ must be ≤ VDD.
2. VREF is expected to equal VDDQ/2 of the transmitting device and to track variations in the
DC level of the same. Peak-to-peak noise (non-common mode) on VREF may not exceed ±1
percent of the DC value. Peak-to-peak AC noise on VREF may not exceed ±2 percent of
VREF(DC). This measurement is to be taken at the nearest VREF bypass capacitor.
3. VTT is not applied directly to the device. VTT is a system supply for signal termination resistors,
is expected to be set equal to VREF, and must track variations in the DC level of VREF.
4. VDDQ tracks with VDD; VDDL tracks with VDD.
5. VSSQ = VSSL = VSS.


/suppress rant about this kind of thing should be in a review. (checking actual voltage vs bios settings.)

 

[dasaint]

makes me wonder if you OC them too badly their chipset goes bad...... or at least thats what i see... maybe they just makin em shittier as they get cheaper....

 

[trudude]

I have gone through 4 different 680i motherboards by different companies. They are all trash. I ended up selling my extra 8800GTX and bought an Asus P5K Deluxe and I have never looked back. Unless nVidia does something spectacular with the 780i's I will never buy another nVidia based board again.

 

[Dan_D]

Maybe I should knock on wood or something.
I have 4x1 GB of Corsair XMS 6400 C4 on my P5N32-SLI board and have not had any problems for 1 solid year.

I have my RAM unlinked and let the board select the voltage.

The PC6400 stuff almost always works better in a 4x1GB configuration than the higher speed modules do. As for you being lucky with your 680i, it happens. Good for you, but many people are having issues with them.

makes me wonder if you OC them too badly their chipset goes bad...... or at least thats what i see... maybe they just makin em shittier as they get cheaper....

I think overclocking quad core CPUs has a dramatic effect on longevity of the 680i SLI boards. I know that any of them I ran with dual core CPUs that weren't DOA are still working and any of them I ran my quad core on overclocked all die in 1-3 months.

 

[Darkatom]

After being through 4 680i boards I finally gave up and went with a Maximus formula instead. I dont have to much faith on the 780i boards but only time will tell.

 

[Dan_D]

After being through 4 680i boards I finally gave up and went with a Maximus formula instead. I dont have to much faith on the 780i boards but only time will tell.

I'm looking forward to the 780i SLI as I'd like a better more reliable solution for SLI than what I've had working with the 680i SLI chipset based boards.

As for the Maximus Formula you've made an excellent choice. It is one of my favorite boards right now.

 

[/dev/null]

Well, I just killed yet another EVGA 680i SLI board.

I was having increasing stability issues, returned the board to stock, that didn't help, so I flashed it with the latest BIOS and now, it's BSOD hell. Swapped the ram, now the thing won't post anymore.

So I'm installing the EVGA Black Pearl as I type this.

The key phrase is "return the board to stock"

Were you overclocking? Does the board work fine at stock or does it die too?

 

[nst6563]

@BillParish - lol... sorry man, didn't mean for you to do all the research I never thought about actually trying to read from a vacant dimm socket. I'm used to seeing the "points" or small contact pads that people commonly read the voltages from. I've seen a lot of that on XS forums.

Thanks tho

:EDIT: I did run across this 1333Mhz fsb pin mod which people seem to be having great success with. I wonder if by "fooling" the chipset into thinking the cpu's default fsb is 1333 that it sets the straps and multipliers accordingly and is somehow "less stressful" on the mb/chipset than simply changing all the settings in bios manualy? Dunno.. might give it a try though. If I'm going to ditch this 680i board in the future...I'd like to go with an x38 board...but right now they're out of my price range

 

[trudude]

680i motherboards will also overvolt and kill your ram. I had to RMA 6 different kits of memory as well.

 

[S[H]ady]

I've been using a 680i board cuppled with a q6600 and 4x1 PC6400.
I have never had any of the issues described here. I've been running this set up since july and i had it overclocked by over 1Ghz 2 days after building it.
I am running 2GB of OCZ system elite, and 2GB of patriot PC6400 (no heat spreaders). The pair have worked flawlessly since the day i put them in.
The only lockup i've had was in the crysis demo but i think that was a driver issue.

Dan_D, terribly sorry for your bad luck. Hope you have better luck with your Black Pearl.

 

[Dan_D]

The key phrase is "return the board to stock"

Were you overclocking? Does the board work fine at stock or does it die too?

I am not sure what you mean by this. Of course I was overclocking. I've never run any motherboard of mine at stock speeds for very long at all. Yes, I admit that many of the premature 680i SLI board deaths I've experienced may be linked with overclocking. However I overclock them as safely as possible. Meaning that I try and make sure everything is rock solid and if it isn't, I back down the settings until it is solid and reliable. Just because you are overclocking doesn't mean that the system will die prematurely. The first board does not behave correctly when everything is set to stock speeds. (To answer your question directly.)

@BillParish - lol... sorry man, didn't mean for you to do all the research I never thought about actually trying to read from a vacant dimm socket. I'm used to seeing the "points" or small contact pads that people commonly read the voltages from. I've seen a lot of that on XS forums.

Thanks tho

:EDIT: I did run across this 1333Mhz fsb pin mod which people seem to be having great success with. I wonder if by "fooling" the chipset into thinking the cpu's default fsb is 1333 that it sets the straps and multipliers accordingly and is somehow "less stressful" on the mb/chipset than simply changing all the settings in bios manualy? Dunno.. might give it a try though. If I'm going to ditch this 680i board in the future...I'd like to go with an x38 board...but right now they're out of my price range

I'll definitely look at the article you linked to.

680i motherboards will also overvolt and kill your ram. I had to RMA 6 different kits of memory as well.

Not necessarily. As I said before I don't believe this is always the case.

I've been using a 680i board cuppled with a q6600 and 4x1 PC6400.
I have never had any of the issues described here. I've been running this set up since july and i had it overclocked by over 1Ghz 2 days after building it.
I am running 2GB of OCZ system elite, and 2GB of patriot PC6400 (no heat spreaders). The pair have worked flawlessly since the day i put them in.
The only lockup i've had was in the crysis demo but i think that was a driver issue.

Dan_D, terribly sorry for your bad luck. Hope you have better luck with your Black Pearl.

Well it sounds like you've had good luck with your motherboard but frankly all of these issues mentioned in this thread by everyone are known issues. The 4x1GB problem is well known as is poor quad core overclocking, and excessive heat problems. I've also seen 680i SLI boards run for six months and then die suddenly. You aren't necessarily off the hook yet, nor is it guaranteed that you will have problems before you upgrade and retire that board/setup from service.

 

[alg7_munif]

790FX boards with AMD Overdrive support look like great enthusiast boards but too bad AMD CPUs and GPUs can't be considered high end anymore. I wanted to go SLi on an Intel CPU but between 680i and X38, I found that the X38 would be a better choice for me.

 

[/dev/null]

I am not sure what you mean by this. Of course I was overclocking. I've never run any motherboard of mine at stock speeds for very long at all. Yes, I admit that many of the premature 680i SLI board deaths I've experienced may be linked with overclocking. However I overclock them as safely as possible. Meaning that I try and make sure everything is rock solid and if it isn't, I back down the settings until it is solid and reliable. Just because you are overclocking doesn't mean that the system will die prematurely. The first board does not behave correctly when everything is set to stock speeds. (To answer your question directly.)

Why not run a test with 2 similar setups: One overclocked, one stock.

Buy everything at the same time (hopefully same batches) & see if one dies.

if the stock one is killing your ram, there is a serious issue.

If only the overclocked one is killing your parts, then perhaps t he 680i sli is just not a good overclocker. I don't think I'd go blaming 680 chipset motherboard manufacturers for dead parts if you are overclocking.

My bike redlines at 15,500. I'd never thinking of changing the rev limiter to 20k and then wondering why other parts that interact with the engine die.

I used to overclock but no longer have time unfortunately, but then again these days I don't see the benefit as much. I no longer do multimedia stuff nor gaming.

 

[Caliche]

if so many ppl are having problems across many manufacture's of 680i's, than why didnt the manufacturing community step-up and do something, at this point, it really dont matter, cuz 780i is here in a week and half, if the news feeds are correct!

I hope 780i isnt the same way, cause that is what i plan on buying so I can do SLI.

Futhermore, how is Nvidia going to encourage or futurely have it in their marketing plan for people to use more than one video card, when they can not get their own chipset to work correctly??!!