Digital Viper-X-
[H]F Junkie
- Joined
- Dec 9, 2000
- Messages
- 15,115
! it scares me, maybe I'll keep that northbridge fan on high, did you ever have any aftermarket nb coolers? like the HR sli one installed?
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Well, I killed another 680i SLI board.
- Thread starter Dan_D
- Start date
! it scares me, maybe I'll keep that northbridge fan on high, did you ever have any aftermarket nb coolers? like the HR sli one installed?
did you ever have any aftermarket nb coolers? like the HR sli one installed?
On the Striker Extrme no. On the eVGA 680i SLI several of them were equipped with a MCW60 waterblock. Even when equipped with a waterblock the last one that failed still ran 20c hotter than the one running with stock air cooling in my other PC.
These threads are so depressing... my eVGA has been good to me, but I have the C00 version, which can't run the quad cores. I'd like to get a new board, I've been thinking Striker and water cooling the CPU/GPU, but with these kinds of posts being so common, it makes me question whether stepping up to a Q6600 is worth the potential hastle. I should just slap water on my current rig on water and call it a day.
My eVGA has been good to me, it's been stable as all hell since the day I bought it... I should just sell one of my vid cards and go X38 with a nice hardware RAID.
Thanks a lot, Dan... you got me all depressed
My eVGA has been good to me, it's been stable as all hell since the day I bought it... I should just sell one of my vid cards and go X38 with a nice hardware RAID.
Thanks a lot, Dan... you got me all depressed
Actually...the C00 CAN run the Q6600's. That's what I'm running right now, STILL waiting on EVGA to send my A1 revision (which they did, and it was a POS so now they're sending me another...only it's a "BR" version...).
I think whether or not your C00 board will run the Q6600's is sort of a crap shoot. Mine seems nice and stable even overclocked the q6600 to 3ghz. However...it won't go 1Mhz more no matter how high I set the bus speed and lower the multiplier (or play with voltages). My friend on the other hand that has the same board and revision, had all kinds of problems running his Q6600 even at stock speeds. He recieved his A1 board and all is fine now.
I'm thinking of Getting the Abit IP35 Pro or the Abit QuadGT X38 board after the holidays and Ebaying this one. I see a LOT of good things about the IP35 Pro, and to be honest...I don't think I'll ever use dual video cards on a system but if the x38 board is close in price I'll get that just to support PCI-E 2.0.
These threads are so depressing... my eVGA has been good to me, but I have the C00 version, which can't run the quad cores. I'd like to get a new board, I've been thinking Striker and water cooling the CPU/GPU, but with these kinds of posts being so common, it makes me question whether stepping up to a Q6600 is worth the potential hastle. I should just slap water on my current rig on water and call it a day.
My eVGA has been good to me, it's been stable as all hell since the day I bought it... I should just sell one of my vid cards and go X38 with a nice hardware RAID.
Thanks a lot, Dan... you got me all depressed
There is no reason to be depressed or even to sell your board if its' working. As I've said several times in this thread already; just because I had problems with mine doesn't mean you will have problems with yours. If it dies it dies, until then enjoy your board.
If you are looking for higher overclocks and you only feel you need one video card, then by all means. Going with X38 is absolutely the way to go.
So if I were going with an x38 board... and was willing to pitch in the ~$70 difference between an IP35 Pro and the x38 board... would you recommend the Abit QuadGT X38 or the Asus Maximus Formula? The only advantage I see to the Abit (maybe it's not really an advantage) is a totally digital PWM power circuit for the CPU where it looks like the Asus is using an 8-phase design.
Looks like both boards are very comparable. Both also use DDR2 so I can use my same memory. I've owned both brands in the past, multiple boards from both actually. And I still use 2 A8n32-SLI Deluxes at home and love them. But then again...I'm still using an Abit BP6 for my firewall box
So if I were going with an x38 board... and was willing to pitch in the ~$70 difference between an IP35 Pro and the x38 board... would you recommend the Abit QuadGT X38 or the Asus Maximus Formula? The only advantage I see to the Abit (maybe it's not really an advantage) is a totally digital PWM power circuit for the CPU where it looks like the Asus is using an 8-phase design.
Looks like both boards are very comparable. Both also use DDR2 so I can use my same memory. I've owned both brands in the past, multiple boards from both actually. And I still use 2 A8n32-SLI Deluxes at home and love them. But then again...I'm still using an Abit BP6 for my firewall box
I haven't heard anything about the abit board yet. The Maximus Formula is simply outstanding and I highly recommend it.
Just got my A1 in from EVGA today to replace my AR for "proper" quad core support. I must say...this A1 board BLOWS ASS. It's not even stable at STOCK speeds...let alone ANYTHING above stock.
Thanks to EVGA and Nvidia, I've made my decision to dump this 680i garbage and go with a P35 or X38 chipset.
I'm sure the 680i is fine with dual cores, but apparently quads just aren't meant for it. I swapped the board and booted - no drivers loaded...ok, fluke...reboot...all the drivers load. I was thinking great, up and running right? wrong. Bsod and reboot...then it hangs at the bios screen. So I reset cmos, get into bios, set it for BELOW stock settings and reboot...flash to P31 bios. Slight improvement...a little more stable at stock speeds. At least no BSODs...just IE closing, games exiting to desktop, crap like that. So I blow away vista and reinstall...same damn problems.
What I did notice though is that at stock...the drivers load...at anything OVER stock...they don't. It boots up and puts me in "punishment mode" with 800x600 res and 16bit color. If I set the speed back to stock or below, windows loads fine (most of the time) but just has problems RUNNING ANYTHING.
so..yeah...whatever...I have enough headaches with my job...don't need them with my PC so Nvidia can kiss my rear on their chipsets. What the hell happened anyway...their stuff was top notch on the AMD side. I'd almost swear I was using a frickin' SIS chipset in here.
Thanks to EVGA and Nvidia, I've made my decision to dump this 680i garbage and go with a P35 or X38 chipset.
I'm sure the 680i is fine with dual cores, but apparently quads just aren't meant for it. I swapped the board and booted - no drivers loaded...ok, fluke...reboot...all the drivers load. I was thinking great, up and running right? wrong. Bsod and reboot...then it hangs at the bios screen. So I reset cmos, get into bios, set it for BELOW stock settings and reboot...flash to P31 bios. Slight improvement...a little more stable at stock speeds. At least no BSODs...just IE closing, games exiting to desktop, crap like that. So I blow away vista and reinstall...same damn problems.
What I did notice though is that at stock...the drivers load...at anything OVER stock...they don't. It boots up and puts me in "punishment mode" with 800x600 res and 16bit color. If I set the speed back to stock or below, windows loads fine (most of the time) but just has problems RUNNING ANYTHING.
so..yeah...whatever...I have enough headaches with my job...don't need them with my PC so Nvidia can kiss my rear on their chipsets. What the hell happened anyway...their stuff was top notch on the AMD side. I'd almost swear I was using a frickin' SIS chipset in here.
My Black Pearl is perfectly stable with my Q6600 even overclocked. I am not sure what the issue is with that board, but I had a similar expereience to the one you are having now. I ended up reinstalling Vista and going back to BIOS P28.
Right now all my issues are solved, and none of my hardware but my motherboard died contrary to what it seemed like before.
a little update as I haven't given up yet....
I think I found the ONLY spot in which the board is stable...
fsb = 1511
multi = 9
1.3v
mem = 756 - 4-4-4-10 1T @ 2.1v
boggles my mind...but it seems stable as a rock so far when it's overclocked to 3.4Ghz... despite having random application crashing at stock speeds...I haven't haven't experienced any of that yet. Ran orthos for a little while, played some Crysis, surfed the web (as IE crashed a lot especially on the [H] with the flash animations) jumped in and out of applications like a monkey on crack. so far so good.
Only bad thing now is that I accidentally cranked a nut too tight on my Gemni II hsf and it sheared the screw off...so now I have to figure out a way to fix that as they were threaded backwards So right now I'm using an Aquagate Viva to cool the cpu...and much to my surprise it's working decently. when running orthos on all 4 cores stressing the cpu, temps never hit above 57c. They were a little cooler with the gemni II, but not by a whole lot. The viva used to cool my old x1800xtpe so I was able to get 800mhz on the core. Since that went to become my son's pc I put the stock cooler back on and had the viva just sitting in a box
I think I found the ONLY spot in which the board is stable...
fsb = 1511
multi = 9
1.3v
mem = 756 - 4-4-4-10 1T @ 2.1v
boggles my mind...but it seems stable as a rock so far when it's overclocked to 3.4Ghz... despite having random application crashing at stock speeds...I haven't haven't experienced any of that yet. Ran orthos for a little while, played some Crysis, surfed the web (as IE crashed a lot especially on the [H] with the flash animations) jumped in and out of applications like a monkey on crack. so far so good.
Only bad thing now is that I accidentally cranked a nut too tight on my Gemni II hsf and it sheared the screw off...so now I have to figure out a way to fix that as they were threaded backwards
So right now I'm using an Aquagate Viva to cool the cpu...and much to my surprise it's working decently. when running orthos on all 4 cores stressing the cpu, temps never hit above 57c. They were a little cooler with the gemni II, but not by a whole lot. The viva used to cool my old x1800xtpe so I was able to get 800mhz on the core. Since that went to become my son's pc I put the stock cooler back on and had the viva just sitting in a box Hey dan, see if you can get a hold of a skulltrail board. That would probably be the best board for you
I had a dual Opteron 254 setup and between the processors, board and 5GB of ram I was out serious cash. No thanks. I can get all the performance I need out of a much cheaper board and one CPU.
my EVGA 680i is 11 months old now, and it never gave me any trouble at all! guess i'm a lucky sob.
x2
However I did kill my 2GB modules after making a mistake when overclocking.
Also my USB on the back is wierd, when I plug my G15 into the very top one and my G5 mouse right next to it, the mouse will not show up when Windows XP boots.
But if I switch the mouse to the USB slot below it, it works great.
Tomorrow I plan on installing Vista x64 and throwing in an additional 2GB of PC2-6400 Crucial Ballistix Tracer and another 8800GT when it arrives. Hopefully this move will be problem free. Heres to hoping anyway....
I have that same USB problem on the board I just swapped out for the Black Pearl. That board also has extremely high NB temps even when being water cooled. No matter how I reapplied the AS5 or Ceramique, I couldn't get the temps down. Even replacing the NB heat sink and fan didn't do the job. I replaced it with an MCW30 waterblock and didn't get much better results. I got a 10c drop in temperature, but even with that drop it was still 20c higher than my other system runs with the stock cooling hardware. On my board though all of my USB ports had the same problem. I couldn't get anything to initialize for a couple of minutes after I saw the Vista logon screen.
I've heard that the USB problems are a sign that your board will die sometime relatively soon. Since yours is still working I'd appreciate it if you'd check something and post the results back here.
I'd like to know what your NB temperatures are. Take readings with any software you want, and let me know what the BIOS says as well. My guess is that your readings will be nearly 20c higher than they should be. If so we might be able to narrow some of the 680i SLI problems down to the 680i silicon itself or possibly to voltage components on the board.
I know one thing, I'm sick of replacing motherboards and If ATI were to come up with something that is 90% or better the power of D9E when its' released I'll gladly jump ship and replace all the NVIDIA hardware with Intel and ATI stuff instead.I'm to the point of sacrificing some performance in order to have something that just works. Or If D9E ends up being twice as powerful as G80 than I may opt to abandon SLI and go with a single card setup. I really hate that idea as my last several machines have all been dual GPU based and most of those were SLI.
are the "normal" A1 boards crazy as hell? I mean, I just got this one in from EVGA to replace the AR...and for the life of me I can't get it to run stable at stock speeds. However, if I bump the fsb by about 15-20mhz it works just fine. Hell, right now I'm running at 3.63Ghz on this Q6600 @ 1.4v (with vdroop mod)...been running that way since about an hour after my last post and I've run Orthos for a while, played some Crysis, surfed the web, watched the screensaver, played some COD4 with an Orthos running in the background....solid as a rock when I overclock the piss out of it...but at stock speeds it's about as stable as a hippo on a highwire. 1066fsb is unstable but 1080fsb is. 1333fsb is unstable (was perfectly stable on the AR) but 1350fsb is. 1600fsb is unstable but 1623 is.
very strange.
very strange.
I have the fan attached to the 680i northbridge heatsink and Everest is telling me the MCP temp is 55º C. CMOS claims the same thing. 131ºF
Here are photos from when I first pulled the heatpipe cooling off and replaced with AS5.
http://picasaweb.google.com/drjester/EVGA680iThermalpasteGoop
Just installed Vista x64 and everything so far is going pretty smoothly.
4GB of Crucial Ballistix Tracer is awesome!
I have the fan attached to the 680i northbridge heatsink and Everest is telling me the MCP temp is 55º C. CMOS claims the same thing. 131ºF
Here are photos from when I first pulled the heatpipe cooling off and replaced with AS5.
http://picasaweb.google.com/drjester/EVGA680iThermalpasteGoop
Just installed Vista x64 and everything so far is going pretty smoothly.
4GB of Crucial Ballistix Tracer is awesome!
Your temps look fine. As I've said before 65c-70c is where I would start to worry. Especially if you get readings like that right after booting the system up cold. I can't say for sure but boards operating around 50c on the NB seem to live longer than the others. Again this leads me to believe there might be a QC issue in regard to 680i SLI MCP silicon but I can't be sure. I don't have enough data to be 100% positive on that.
Your temperature looks good based on my experience.
A friend of mine was having USB issues with his eVGA 680i SLI board tonight while I was at his house. I had him check his BIOS and his north bridge temperature was in the 80c range. I suspect his will die very shortly. When that board was new it ran within the normal temperature range, so I don't know if the problem I've noticed with the NB temps being to high is something that the boards that fail early have from the start or if it is a gradual condition that hits these boards as they age. My friend's system isn't overclocked and is running good parts and decent case cooling, but the NB temps are now out of control. The system seems to work ok outside of USB flakyness right now, but that's the same thing my last 680i SLI board did prior to death. Incidently my buddy is running a dual core E6600 and not a quad core processor.
Granted 13 or 14 eVGA boards I've dealt with, (owning 11 personally) isn't necessarily a large enough sampling to determine what the exact cause of some of these issues are, but it is a start.
Keep the information coming and maybe we can narrow down the cause of these issues.
mystical2008
n00b
- Joined
- Jan 26, 2008
- Messages
- 4
I am a seasoned computer tech and I am troubled by the amount of problems many are having with this 680i motherboard. I recently last year built my friend a brand new system with the latest hardware. We now have to do a fourth RMa on the same mobo due to double dashes on the board(flatline as i call it).
I do not know if it is an XFX problem as well, but it happens to be on EVGA boards. Nvidia does make the chipset and as such there could be design flaws with the northbridge. Many times companies pass the buck to others but ultimately someone has to fess up ot maybe cover their tracks.
I did notice on the new 780i mobos that they have a cooling fan over the northbridge heat sink. Now this obviously is a slight improvement over the 680i. I know many of us have to buy after market products to keep the temps cool in the system but can't these manufacturers instead of taking the easy way out, provide all you need in the boards and contents that come with them. It is amazing that RMA's from EVGA, they send out a refurbished mobo and the one we received was definitely used as we saw that someone had removed previously a thermaltake adhesive and bracket on the backside of the mobo. Within an hour of hook up and running the computer froze and locked up. The mobo went "--" again and thus the cycle of repeatedness begins.
I feel that research by the company needs to be done and that they should aloow users who have documented problems with the 680i upgrade to the 780i for free to help compensate for the downtime many suffer. Not all of us have big money to build 2-5 rigs.
Dan is right I think in the northbridge is the culprit. If you check EVGA site as they tout the new 780i, you will see a fan on the northbridge made by nvidia.
Comspiracy possibly or it jsut a way for nvidia to cover the flaws in their design and hush it up and pass the buck onward.
It boils down to obviously better quality control and more thorough testing by manufacturers before releasing their products. I know many people have damaged boards on their own I am sure but a majority of us take care of our rigs since we put our heart and soul into building them.
Dan you hit the nail on the head with the northbridge as I am in agreement. Hopefully, the companies are listening to the concerns and not jsut releasing a board to cover up the flaws in the previous. Maybe this new board addresses those issues and fianlly lays to rest the problems so many people have been having.
I do not know if it is an XFX problem as well, but it happens to be on EVGA boards. Nvidia does make the chipset and as such there could be design flaws with the northbridge. Many times companies pass the buck to others but ultimately someone has to fess up ot maybe cover their tracks.
I did notice on the new 780i mobos that they have a cooling fan over the northbridge heat sink. Now this obviously is a slight improvement over the 680i. I know many of us have to buy after market products to keep the temps cool in the system but can't these manufacturers instead of taking the easy way out, provide all you need in the boards and contents that come with them. It is amazing that RMA's from EVGA, they send out a refurbished mobo and the one we received was definitely used as we saw that someone had removed previously a thermaltake adhesive and bracket on the backside of the mobo. Within an hour of hook up and running the computer froze and locked up. The mobo went "--" again and thus the cycle of repeatedness begins.
I feel that research by the company needs to be done and that they should aloow users who have documented problems with the 680i upgrade to the 780i for free to help compensate for the downtime many suffer. Not all of us have big money to build 2-5 rigs.
Dan is right I think in the northbridge is the culprit. If you check EVGA site as they tout the new 780i, you will see a fan on the northbridge made by nvidia.
Comspiracy possibly or it jsut a way for nvidia to cover the flaws in their design and hush it up and pass the buck onward.
It boils down to obviously better quality control and more thorough testing by manufacturers before releasing their products. I know many people have damaged boards on their own I am sure but a majority of us take care of our rigs since we put our heart and soul into building them.
Dan you hit the nail on the head with the northbridge as I am in agreement. Hopefully, the companies are listening to the concerns and not jsut releasing a board to cover up the flaws in the previous. Maybe this new board addresses those issues and fianlly lays to rest the problems so many people have been having.
All the 680i SLI reference boards sold by EVGA, XFX, BFG, ECS, and possibly other companies are all the same. These boards are built by Foxconn and contracted to do so by NVIDIA. NVIDIA then sells the boards to EVGA, BFG and others and those boards are stickered with the brand name and labels corresponding to each company.
The 680i SLI reference boards I have seen all have fans on the north bridge. The ASUS and some other non-reference design boards don't necessarily. Even so the ASUS board for example runs cooler than the reference board does.
Actually, they do. The problem is that many machines are operated by users who are in ignorance of the operations guidelines set by the manufacturer. Most people's homes probably provide ambient temperatures that are less than ideal for hot equipment like this. If it is 75F in your house it is probably a bit on the warm side for your computer. Also many people shove their computers out of the way inside cabinents and other places where airflow is less than ideal. Also many people who live in older homes or apartments may have electical circuits that aren't up to the task of running so much modern equipment. I am not saying this is true of every person who has had issues with the 680i SLI boards, but it is something to be considered. I don't qualify for any of the problems I just mentioned and I've killed 10 of these things.
I understand where you are coming from here, but I'm going to say bullshit here. First off compensation for downtime on a computer used for recreation is stupid. If your computer is mission crticial and used for business then you should have purchased a Dell or HP workstation with next business day or 4 hour response time for parts and technicians to maintain such critical equipment. Anything else is foolish. Also if you read the warranty statements and EULA's from hardware and software companies you will find that these companies are quite well protected from such liabilities. Lost wages or compensation for your time is YOUR problem. I hate to sound like an asshole or a bureaucrat but let's be realistic. You need to use the right tool for the right job and a system you built yourself that doesn't have the proper spare parts on hand or any old consumer unit without proper coverage is not the correct tool for the job. Additionally if you want to depend on a system for work, that system should be built on an Intel chipset and not an NVIDIA chipset. Hardware 101 and common sense should tell you than Intel chipsets are SUPERIOR to all others where Intel processors are concerned. No they aren't SLI compatible but that shouldn't be a problem for a work computer.
I do not think that the north bridge is to blame by itself. The problem is with the reference design which uses an outdated 6-phase power design and generally speaking regulates voltages very poorly. Poor BIOS and other design flaws may also be contributing factors to premature 680i SLI death. Additionally it seems evident to me that the voltage design is ok for dual cores but not quad core processors. The boards running quad core processors seem to die much faster than those running dual core chips.As for the 780i SLI board, it still seems to be lacking in a number of areas. The north bridge fan alone isn't what's needed to make it a better board. The 780i SLI reference boards are still largely identical to the design of the 680i SLI boards. (Which as I said earlier also have NB fans.) All indications are that the 780i SLI reference boards are as bad as the 680i SLI boards and in some cases worse.
I'm not sure what you mean here.
To a degree it does. However I think the design is flawed when it comes to powere circuitry and design. In fact it would seem that EVGA themselves have validated this thinking with their FTW Edition. What they did was upgrade the power design from 6 to 8 phase, add newer higher quality solid state capacitors and upgrade the PWMs. Indeed that would seem to indicate a deficiency in power system design. Otherwise why would it need such drastic improvement?
Dan you hit the nail on the head with the northbridge as I am in agreement. Hopefully, the companies are listening to the concerns and not jsut releasing a board to cover up the flaws in the previous. Maybe this new board addresses those issues and fianlly lays to rest the problems so many people have been having.[/QUOTE]
Again it isn't the north bridge silicon that is to blame. If it is part of the equation it would be due to quality control but as I've stated I think the power circuitry is really to blame. Indeed improper voltage regulation can be responsible for the north bridges freaking out and literally baking themselves. The 680i SLI north bridge silicon not being at fault can be evidenced by issues found with the non-reference boards. The non-reference boards are often more aggressively tuned in regard to memory timings and in the way they handle SPD values differently. OCZ has much information on this and even released their own version of the P5N32-E SLI and Striker Extreme BIOS' to correct for their improper memory settings. Those boards also have a lower failure rate from what I can tell and when the memory settings fall in line to where they need to be or the memory chosen works well with those particular boards they often achieve greater stability than the reference boards do.
Just my $.02.
Ihaveworms
Ukfay Ancerkay
- Joined
- Jul 25, 2006
- Messages
- 4,628
So far, my board is about 3 months old and it just doesn't like going above 3ghz with my Q6600. I was reading about the MCP temps and I redid the thermal paste and my MCP still stays around 59-61. Is that ok?