I've got a set of corsair 8500c5d memory that has been working well so far. I'm running it in an evga 680i and after flashing a bios I get the c1 post error if the voltage hasn't been manually set. To get around this I end up using just one stick to make it boot. However, the odd thing is that only one of my two sticks will boot at the default voltage(1.85v?). The other stick refuses to boot until it recieves 2.25v or greater. After that it runs completely fine. When this memory was new I don't recall having a problem booting them both at default voltage, until I could get into bios to change them to 2.2v. Any ideas why this one stick now wants more power to operate? The other one doesn't exhibit this issue. Both are memtest stable.
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Corsair 8500c5d voltage question
- Thread starter Tigertank
- Start date
Redbeard
Official Corsair Rep.
- Joined
- Nov 3, 2005
- Messages
- 1,859
The 680i chipset works kind of wonky with the single module in it. Usually it boots at 667 or 800 MHz at very relaxed latency with one module, even if SLI memory is enabled.
Which BIOS are you using, and is SLI ready memory enabled?
I know for a fact that those modules won't do 1066 MHz @ 1.85volts, you're going to have to set the voltage to 2.2V one way or another, either through SLI ready memory or manually setting everything.
The default timings it uses if you clear the cmos are 800mhz 5-5-5-18 1.85v. That's what it does in my case anyway. Sli ready memory seems to be disabled by default as well. Typically I run it without sli ready memory enabled and just set everything manually. The bios I'm using is p24.
Now upon doing some further testing, if I enable sli ready memory, have the voltage manually set to 2.2v, then the module that will normally not post on its own under default settings will post. However, if I try to set things manually, with sli ready memory disabled, to something conservative like 800mhz 5-5-5-15 2.2v then that module in question will not post if it is the only module in the system. The "good" module will post under any normal conditions with or without the second module. (cmos defaults, manual settings @ 2.2v etc.)
The reason I am curious as to what is going on here, is that as of late I've been experiencing alot of bsod's with the nv4_disp.dll. I've narrowed down the possibilities of the issue by testing different hardware configurations. The symptoms are reproduceable with two different video cards. (8800gtx and 7900gt) Both of which function without issue if put into another system. So I ruled that out as the cause. The memory will pass memtest86 indefinately. The longest session I ran after this issue became apparant, was slightly over 24 hours with no errors. The only odd ball issue with the memory is the post issues I mentioned above. The whole system had been completely rock solid stable for the first 60 days I had owned the parts.
I have since rma'd the 680i since I believed it to be the cause of all of this. The rma board was completely stable from mid day friday until sunday evening. Whenever I wasn't using the system I was either running orthos or looping 3dmark for stability testing. Orthos has never returned an error on either large or small fft's. And 3dmark ran fine for the first 48 hours. Now the bottom has fallen out of the bucket again, so to speak. Back to bsod's when using 3d applications and even while web browsing the system can become unresponsive and return a nv4_disp bsod. One thing to note is that these problems tend to take more time to occur if only one module of memory is in the system. But the bsod can occur on either module if tested alone. The bsod also occurs at stock or oc'd configurations. Hell, even underclocked and really lax timings.
I find it hard to believe that two modules would exhibit problems at exactly the same time. That's why I wanted to get an expert opinion about it. At this point I really want to point the finger at the 680i again, but EVGA's tech's are claiming this to usually be a memory issue. And while I would tend to agree that bsod's typically are memory related one way or another, a system doesn't just magically work under stress for 48 hours then start passing out bsod's like candy if it has problems with the memory modules.
What do you think Redbeard? Thanks in advance for taking time to look at this, I certainly appreciate it.
Now upon doing some further testing, if I enable sli ready memory, have the voltage manually set to 2.2v, then the module that will normally not post on its own under default settings will post. However, if I try to set things manually, with sli ready memory disabled, to something conservative like 800mhz 5-5-5-15 2.2v then that module in question will not post if it is the only module in the system. The "good" module will post under any normal conditions with or without the second module. (cmos defaults, manual settings @ 2.2v etc.)
The reason I am curious as to what is going on here, is that as of late I've been experiencing alot of bsod's with the nv4_disp.dll. I've narrowed down the possibilities of the issue by testing different hardware configurations. The symptoms are reproduceable with two different video cards. (8800gtx and 7900gt) Both of which function without issue if put into another system. So I ruled that out as the cause. The memory will pass memtest86 indefinately. The longest session I ran after this issue became apparant, was slightly over 24 hours with no errors. The only odd ball issue with the memory is the post issues I mentioned above. The whole system had been completely rock solid stable for the first 60 days I had owned the parts.
I have since rma'd the 680i since I believed it to be the cause of all of this. The rma board was completely stable from mid day friday until sunday evening. Whenever I wasn't using the system I was either running orthos or looping 3dmark for stability testing. Orthos has never returned an error on either large or small fft's. And 3dmark ran fine for the first 48 hours. Now the bottom has fallen out of the bucket again, so to speak. Back to bsod's when using 3d applications and even while web browsing the system can become unresponsive and return a nv4_disp bsod. One thing to note is that these problems tend to take more time to occur if only one module of memory is in the system. But the bsod can occur on either module if tested alone. The bsod also occurs at stock or oc'd configurations. Hell, even underclocked and really lax timings.
I find it hard to believe that two modules would exhibit problems at exactly the same time. That's why I wanted to get an expert opinion about it. At this point I really want to point the finger at the 680i again, but EVGA's tech's are claiming this to usually be a memory issue. And while I would tend to agree that bsod's typically are memory related one way or another, a system doesn't just magically work under stress for 48 hours then start passing out bsod's like candy if it has problems with the memory modules.
What do you think Redbeard? Thanks in advance for taking time to look at this, I certainly appreciate it.
Redbeard
Official Corsair Rep.
- Joined
- Nov 3, 2005
- Messages
- 1,859
Well, if you've used two different video cards, and those cards work fine on other systems, it's probably not them.
If you've had the motherboard RMAd, and it still didn't help, it's probably not the motherboard.
The memory passes memtest and runs orthos and 3DMark stable, I'd like to say that it's not the memory, either.
This leaves me with a few suspicions. One of which is software. I don't know if you mentioned if you've formatted a fresh hard drive and loaded up new drivers, but that would be my next suggestion. If for nothing else, just to rule out software as a cause.
In my experience, I've rarely seen bad memory cause the nv4_disp.dll blue screen, but I'm sure now that I've said that somebody else here will jump in and say it happens all the time.
My other suggestion is to try the memory on a different system.
If you leave the voltage set to auto and enable SLI Ready Memory, it should automatically set the voltage, as well.
Just so I make sure I'm on top of things: Manually setting the memory to 1066 MHz @ 5-5-5-15-2T @ 2.2V works fine in memtest and orthos, but the blue screen still pops up in 3D apps?
At this point, if it's not the motherboard, memory, or video card, and the software reinstall doesn't help, the next step is to start looking at weird stuff like the processor, power supply, SATA cables, etc.
I'd gladly RMA the memory for you if I thought it would help, but I have a feeling we'd send you a new set and the problem would still be going on, because I doubt it's memory based.
When you say 3D apps, which programs specifically are you talking about? I'd like to try and recreate your bugs in the lab.
Yeah the system has recieved a format and fresh drivers once on each of the boards just to be on the safe side.
One thing about voltage, I just got off the phone with one of the EVGA tech's and he claimed that with p24 they had seen some cases where the board was undervolting to the memory and the northbridge. Whether or not that is true I don't know, but on the last board I had rolled back all the way to p21 and could reproduce the problem on p21,p23, and p24. Out of curiosity, is there a safe place you can get a reading with a multimeter off a dimm slot without risking damage? I've even pushed 2.3v through it at 800mhz with bsod's, so that sounds like a dubious cause to me. But I'll test with different voltages within reason.
And yeah, I really doubt it's the memory too. Just because of the results of all the testing I've done and the fact that both sticks have the problem when seperated. This is actually a pair you swapped for 6400c4d's, so unless they came out of that bag of evil memory you keep under your desk, I'm sure they're fine. I'll see what I can do about testing the memory in another system. I've not done that yet just because my other system is an amd ddr board, but I know someone who I think has a lower end c2d ddr2 board sitting around that I can test with.
As for the frequency and timings, yes for testing purposes I have been using a manual setting of 1066mhz 5-5-5-15 2.2v and have played around by even dropping it to 800mhz and bumping it as high as 2.3v. With the exception of enabling sli memory to make that one odd ball module post by itself for testing. It will run 1066 4-4-4-12 and pass memtest and orthos all day long, but I've refrained from using aggressive timings until I can pin point this problem.
3d applications are what bring about the bsod the quickest. However, it's possible to get it web browsing as well. Typically a website with something animated has caused it. The very first time I had gotten one of these bsod's was actually when I let the system idle over night and came back to it the next morning and opened a web page. The system becomes unresponsive for 3-5 seconds then makes an attempt to jump back to what it was doing then usually locks up completely followed by a nv4_disp.dll bsod.
The applications that can cause a bsod are rather extensive. I've experienced it with 3dmark regularly, though this board was stable the first 48 hours. I've also encountered it with hl2 and other source based games, battlefield 2 & 2142, rainbow six: vegas, and even app's like 3d studio max will bsod it. Let's just say I havn't found any graphically intensive application that hasn't resulted in a problem. I really had cause to believe the 8800gtx was failing when all this started happening, but I tested it in my amd system extensively without missing a beat. And it's known working 7900gt bsod'd in this system too. So I'm stumped there.
I've also tried to rule out the remaining hardware like the psu as well. I've monitored and logged the voltage read outs when these bsod's occur and everything is very flat and stable without voltages jumping around etc. I'll probably try another psu on this hardware to see what happens. Even the little things like the cables have been checked, and rechecked to no avail.
There has been talk of other users having similar issues on the evga forums but no definate fix. Some got it from video card failure, some from out of date directx packages, and one other case that I know of where the user experienced short term stability after rma'ing the board then eventually back to regular bsod's. Obscure problems are such a pain in the ass.
I may split my two disks out of raid 0 and have one with xp installed and one with vista this week, as I was planning to upgrade anyway. Maybe something differently will happen with vista, I don't know. Worst case is a bsod.
One thing about voltage, I just got off the phone with one of the EVGA tech's and he claimed that with p24 they had seen some cases where the board was undervolting to the memory and the northbridge. Whether or not that is true I don't know, but on the last board I had rolled back all the way to p21 and could reproduce the problem on p21,p23, and p24. Out of curiosity, is there a safe place you can get a reading with a multimeter off a dimm slot without risking damage? I've even pushed 2.3v through it at 800mhz with bsod's, so that sounds like a dubious cause to me. But I'll test with different voltages within reason.
And yeah, I really doubt it's the memory too. Just because of the results of all the testing I've done and the fact that both sticks have the problem when seperated. This is actually a pair you swapped for 6400c4d's, so unless they came out of that bag of evil memory you keep under your desk, I'm sure they're fine. I'll see what I can do about testing the memory in another system. I've not done that yet just because my other system is an amd ddr board, but I know someone who I think has a lower end c2d ddr2 board sitting around that I can test with.
As for the frequency and timings, yes for testing purposes I have been using a manual setting of 1066mhz 5-5-5-15 2.2v and have played around by even dropping it to 800mhz and bumping it as high as 2.3v. With the exception of enabling sli memory to make that one odd ball module post by itself for testing. It will run 1066 4-4-4-12 and pass memtest and orthos all day long, but I've refrained from using aggressive timings until I can pin point this problem.
3d applications are what bring about the bsod the quickest. However, it's possible to get it web browsing as well. Typically a website with something animated has caused it. The very first time I had gotten one of these bsod's was actually when I let the system idle over night and came back to it the next morning and opened a web page. The system becomes unresponsive for 3-5 seconds then makes an attempt to jump back to what it was doing then usually locks up completely followed by a nv4_disp.dll bsod.
The applications that can cause a bsod are rather extensive. I've experienced it with 3dmark regularly, though this board was stable the first 48 hours. I've also encountered it with hl2 and other source based games, battlefield 2 & 2142, rainbow six: vegas, and even app's like 3d studio max will bsod it. Let's just say I havn't found any graphically intensive application that hasn't resulted in a problem. I really had cause to believe the 8800gtx was failing when all this started happening, but I tested it in my amd system extensively without missing a beat. And it's known working 7900gt bsod'd in this system too. So I'm stumped there.
I've also tried to rule out the remaining hardware like the psu as well. I've monitored and logged the voltage read outs when these bsod's occur and everything is very flat and stable without voltages jumping around etc. I'll probably try another psu on this hardware to see what happens. Even the little things like the cables have been checked, and rechecked to no avail.
There has been talk of other users having similar issues on the evga forums but no definate fix. Some got it from video card failure, some from out of date directx packages, and one other case that I know of where the user experienced short term stability after rma'ing the board then eventually back to regular bsod's. Obscure problems are such a pain in the ass.
I may split my two disks out of raid 0 and have one with xp installed and one with vista this week, as I was planning to upgrade anyway. Maybe something differently will happen with vista, I don't know. Worst case is a bsod.
Another symptom I've started to experience with this system, is that when in bios I have had the system freeze on me while changing some settings. I was able to reproduce it several times just by letting the system idle in bios for a few minutes and it would become unresponsive. Is the system heavily dependant on memory while in bios? I think that also suggests that it's definately not a software problem if it hangs in bios. I've since pulled the 8800gtx and strapped in an ancient s3 trio64 pci vga card with a massive 1mb of vram, just to cut it out of the equation and will see if I can reproduce this. I'm also going to take a really close look at any voltage variation between having the 8800gtx in or out of the system. My psu has three 19a 12v rails I believe. Which should be sufficient, but is worth taking an extremely close look at. I'm sure windows would never bsod without having the nvidia drivers being utilized, but hopefully I can find a clear hardware problem to reproduce.
One other thing I thought I'd bring to attention is this thread:
http://http://www.evga.com/community/messageboard/topic.asp?TOPIC_ID=25544&whichpage=2
It seems other users are experiencing the same thing that I am. Have you guys experienced anything similar in house Redbeard? I know that you use the 680i to qualify some of your modules, but have you ever had any in a test platform for an extended period of time that went bad? It seems really really odd that this occured multiple times for different people on evga's 680i all operating around 4-6 weeks prior to some sort of failure somewhere.
http://http://www.evga.com/community/messageboard/topic.asp?TOPIC_ID=25544&whichpage=2
It seems other users are experiencing the same thing that I am. Have you guys experienced anything similar in house Redbeard? I know that you use the 680i to qualify some of your modules, but have you ever had any in a test platform for an extended period of time that went bad? It seems really really odd that this occured multiple times for different people on evga's 680i all operating around 4-6 weeks prior to some sort of failure somewhere.
Redbeard
Official Corsair Rep.
- Joined
- Nov 3, 2005
- Messages
- 1,859
We use this board extensively for qualifying memory because it's the fastest. My main job (besides posting here) is to be an Apps Specialist which means I try to find out new applications for various hardware (basically, creating new products and pitching them to the higher-ups). I've got a lot of eVGAs in the lab, within 20 feet of me I think there are half a dozen right now, with the P24 BIOS. My boss has one in his home system and has for about 2 months now, hasn't seen it fail.
I know we've had one board that used to run exceptionally fast (1200 MHz+ on memory) that no longer does it reliably, but that board seems to work fine at 800 MHz, so I'm not really sure what the deal is with that one.
This is a very bizarre issue. Those two sticks of RAM are identical, same chips, same SPD, same PCB, same components. It's just crazy that one boots and the other doesn't. The BIOS, I imagine, is the culprit, since you seemed to not have the issue until you flashed it. That's all I can think of.
I joined this forum just to post in this thread.
I started with evga 680i + 9136C5D. Everything worked fine for 1-2 weeks, then it wouldn't boot, post, or gave BSOD with nv4 disp or usb driver etc. BUT - it would boot with one of the sticks.
Received 6400C3DF as replacement, and did the same thing unfortunately - worked fine at first then became defective. Again, one stick would work, one wouldn't. Then I put my old 9136 back in, and each stick would work individually. WEIRD!
Anyway, RMA'd the mobo and got a new one, along with some 8500C5D - worked fine for 2 weeks. Now, same problem again! BSOD, can't post, crashes etc. No idea what causes it and it's running at stock the whole time (except I manually set voltage to 2.2v just in case).
It seems like the mobo does something to the ram... which causes it to break over time.
I started with evga 680i + 9136C5D. Everything worked fine for 1-2 weeks, then it wouldn't boot, post, or gave BSOD with nv4 disp or usb driver etc. BUT - it would boot with one of the sticks.
Received 6400C3DF as replacement, and did the same thing unfortunately - worked fine at first then became defective. Again, one stick would work, one wouldn't. Then I put my old 9136 back in, and each stick would work individually. WEIRD!
Anyway, RMA'd the mobo and got a new one, along with some 8500C5D - worked fine for 2 weeks. Now, same problem again! BSOD, can't post, crashes etc. No idea what causes it and it's running at stock the whole time (except I manually set voltage to 2.2v just in case).
It seems like the mobo does something to the ram... which causes it to break over time.
Redbeard
Official Corsair Rep.
- Joined
- Nov 3, 2005
- Messages
- 1,859
To test this for you guys, I'm going to build a system up and do a long-term test. I'll use the system daily for a month or so with 8500C5s.
Out of curiousity, could you guys both get me your info on the BIOS revision, which processor you're using, and the voltages to your northbridge, CPU, etc?
Original setup:
E6600 @ various speeds/voltages
All other voltages were default except memory (corsair default set manually).
Fan installed on the chipset and I have corsair memory fan installed
BIOS was P23beta I believe. I updated the bios first thing because of my X-fi
Current:
E6600 @ various speeds/voltages
All voltages were default except memory (corsair default set manually)
Fan installed on the chipset and I have corsair memory fan installed
BIOS was something (didn't check), tried updating to P24 to fix problem but it remains
Full system:
EVGA 680i
E6600
8800GTX SLi
X-fi Music
PCP&C 1kW-SR
Benq DVD-RW (This had problems with NF4 IDE drivers I believe, so I didn't install them for the 680i either. It worked fine with my DFI SLI-DR as long as drivers were not installed)
9136C5D/6400C3DF/8500C5D with memory fan
I just wanted to add:
I had done things like play Oblivion for several hours straight, also Prey etc. Ran Prime95 for hours with 0 errors. Then I'd wake up one day, start to use the computer to email/websurf and it'd freeze. From that point it's just endless crashing. Right now I'm using the computer but it randomly freezes.
Well my original board that started all of this bsod business was running the following:
p23 Bios (when the problem presented itself, upgraded to p24 when it was available)
e6600 @ 3.6ghz 1600x9, 1.40vcore (1.45v in bios before vdroop)
cpu fsb 1.4v
vdimm 2.3v 1066mhz 4-4-4-12 2t
spp northbridge 1.45v
mcp southbridge 1.50v
The subtimings were run at 3-22-5-9, 7.8us for quite awhile,
rest was defaults mostly, with the exception of disabiling all spread spectrums etc.
The new rma board is the same with the exception of increased voltages to the nb and sb to try to stabilize it. The symptoms continue at 1.5v, and I wasn't going past that. Running the cpu and fsb stock w/ increased voltages doesn't help either.
Another problem that just happened today is that now I can't even make the board post with my 8500's. The stick that was problem free now just gets stuck at 1D post code and won't post after clearing the cmos. The other stick continues to C1 as it has been doing for some time now. So either the board finally cooked the memory or the board gave up the ghost.
I'm going to get some el cheapo ddr2 and throw it in to see what happens. I need some testing modules anyway. If I can't get these 8500's working and the board posts with other ram, I'd be willing to send them to you Redbeard rather than straight rma if you think it's possible to diagnose exactly what happened to the modules? The fact everyone with this problem is not getting any memtest errors just compounds the mystery. It may turn out there's not a thing wrong with them and the boards are just going to hell on me for one reason or another. At this point it's nothing but problems though.
p23 Bios (when the problem presented itself, upgraded to p24 when it was available)
e6600 @ 3.6ghz 1600x9, 1.40vcore (1.45v in bios before vdroop)
cpu fsb 1.4v
vdimm 2.3v 1066mhz 4-4-4-12 2t
spp northbridge 1.45v
mcp southbridge 1.50v
The subtimings were run at 3-22-5-9, 7.8us for quite awhile,
rest was defaults mostly, with the exception of disabiling all spread spectrums etc.
The new rma board is the same with the exception of increased voltages to the nb and sb to try to stabilize it. The symptoms continue at 1.5v, and I wasn't going past that. Running the cpu and fsb stock w/ increased voltages doesn't help either.
Another problem that just happened today is that now I can't even make the board post with my 8500's. The stick that was problem free now just gets stuck at 1D post code and won't post after clearing the cmos. The other stick continues to C1 as it has been doing for some time now. So either the board finally cooked the memory or the board gave up the ghost.
I'm going to get some el cheapo ddr2 and throw it in to see what happens. I need some testing modules anyway. If I can't get these 8500's working and the board posts with other ram, I'd be willing to send them to you Redbeard rather than straight rma if you think it's possible to diagnose exactly what happened to the modules? The fact everyone with this problem is not getting any memtest errors just compounds the mystery. It may turn out there's not a thing wrong with them and the boards are just going to hell on me for one reason or another. At this point it's nothing but problems though.
Redbeard
Official Corsair Rep.
- Joined
- Nov 3, 2005
- Messages
- 1,859
At this point I'm not convinced that RMAing the modules would help, as I'm starting to suspect a memory controller issue. I'll gladly set up an RMA for you, if you want to go that way, but I'll wager if you send them in they'll pass whatever tests I can run.
Like I said, if you want to do an RMA, feel free to PM me, but I can't guarantee it will help.
AuroraProject
Gawd
- Joined
- Dec 13, 2004
- Messages
- 820
Hmm, I have the Dominator 1250's in a 680i now and it's running fine (at 1250mhz). The build is only 4 days old though, I haven't had any problems so far. <knocks on wood>
Redbeard
Official Corsair Rep.
- Joined
- Nov 3, 2005
- Messages
- 1,859
And the other 680i SLI boards we use have been running 1200+ for the past 2-3 months without any problems, so maybe it's just that one board.
I agree that it's most likely not the modules. The fact that this rma board got stressed really hard for 2 days straight would have turned up a crash somewhere if the modules were at fault. I just suggested that, thinking that you would have the equipment at your disposal to make 100% sure absolutely nothing is going on with the modules. I don't really want them replaced if they are working normally. I intend to exhaust every other possibility with this board before going that route though.
I'm gonna talk with evga and get yet another board from them. Third time is a charm? Let's hope so. Either I got extremely unlucky or the problem is more wide spread than I think it is.
I'm gonna talk with evga and get yet another board from them. Third time is a charm? Let's hope so. Either I got extremely unlucky or the problem is more wide spread than I think it is.
Tornadojoe
n00b
- Joined
- Jan 18, 2007
- Messages
- 9
Sure glad you guys are on this board ,I`m Running eVGA 680I with 8500c5d PCP&C 1kw no problems yet other then 25 post code for 5 mins cleared Bios still no help,Hit space bar,Got my Video back been fine so far.Been 3 days hope for the best.I did remove the putty off the chipset and reset it with # 5 Back in the day i could buy 3 different Mobo`s if one had problems throw the other board in,at 250-400 each cant do that anymore.This board cant be that bad ? or can it.why the putty on the chipset.They run out of Mud ...
I contacted EVGA again, and this time the tech didn't bite on an rma as the solution. He claimed they had some systems reproducing this issue as well, and will be getting in touch with Nvidia to push for a bios release or board revision to solve this. Guess I'll just have to sit and wait to see what Nvidia's answer is. Atleast it's not the obscure little problem nobody has heard of anymore.
Thermal putty isn't near as efficient for heat transfer as arctic silver or any other thermal compound. The only thing it is better at is filling larger gaps of space where components don't seat flush on a heatsink. The putty solution was just a cost cutting measure Foxconn used for mass production. I would have done it differently, but to each their own. For applications like the NB and SB where the heatsink contact is good it doesn't make any sense to have a massive glob of putty caked all over the chip and anything near it.
Man, I just ordered this memory from newegg yesterday....figured it would go nice on the 680i board....
It was even posted on evga's website as "compatible memory" - thats why i ordered it.
Haha, now i read this post...wow I have bad luck when it comes to ordering parts. Last week I got the 680i with the kentsfield qx6700 . I set it all up to learn that the 680i is an out of date chipset that cant overclock the new quad-cores...wow.
Now this! so you're saying im most likely to see the video blank out several times a day with the Corsair memory? wow...and spending $370 bucks on these suckers gets you nowhere
It was even posted on evga's website as "compatible memory" - thats why i ordered it.
Haha, now i read this post...wow I have bad luck when it comes to ordering parts. Last week I got the 680i with the kentsfield qx6700 . I set it all up to learn that the 680i is an out of date chipset that cant overclock the new quad-cores...wow.
Now this! so you're saying im most likely to see the video blank out several times a day with the Corsair memory? wow...and spending $370 bucks on these suckers gets you nowhere
Redbeard
Official Corsair Rep.
- Joined
- Nov 3, 2005
- Messages
- 1,859
It's not really a hugely widespread problem, it's certainly not 100%. From the amount of times I've seen it mentioned, I believe a few things:
1) It is a real problem that needs to be addressed, but it doesn't seem to be memory based.
2) It's a very, very small percentage of the population that is seeing this issue.
3) It's extraordinarily difficult to track down what the issue actually is because it takes long periods of time to duplicate.
My first theory is that nv4_disp.dll error has something to do with the video drivers, and that makes pretty good sense. It's a distinct possibility that it's related to some obscure software program or piece of hardware or revision of hardware that all sufferers share.
I can guarantee you it's not happening to a large percentage of our customers, though. The XtremeSystems thread seems to reinforce the fact that other peoples' systems are suffering this problem, and many of them aren't even using Corsair.
Redbeard, I don't know much about technology but to me it seems that Ram either works or it doesn't - so I don't think the RAM is the problem.
I don't believe it's necesarily a software issue either because I had many POSTing issues (Memory detect errors, Code 25, Code C1, Code C3). It would hang at all these places at different times. I was also getting mostly usb BSOD (not nv4_disp).
Maybe there is simply an large error in the 680i chipset architecture or something?
I'm glad someone is trying to address is because I've read of a few instances with people having similar problems.
I don't believe it's necesarily a software issue either because I had many POSTing issues (Memory detect errors, Code 25, Code C1, Code C3). It would hang at all these places at different times. I was also getting mostly usb BSOD (not nv4_disp).
Maybe there is simply an large error in the 680i chipset architecture or something?
I'm glad someone is trying to address is because I've read of a few instances with people having similar problems.
I am also running into this issue.
my specs are:
680i P24
2x1GB 8500C5D
2 x asus 8800GTX SLI
silverstone 850w SLI PS
1 x raptor 36GB
X-Fi fatality pro gamer
When i first noticed the issues I would get the C1 error and then have to jockey memory slots configs to geet it to boot, then I went to one mem stick and would get it to boot and set all timings and voltages correctly for the ram.
i have tried all configs already suggested with no luck. i am currently awaiting an RMA board from eVGA.
Since I have bought some crucial 6400 ram and have been running both sticks with out a problem at all...
I dont think that it is the ram, I think that it is the chipset or mem controller...
it doesn't like the corsair for some reason?
although the IC's on my crucial are the same IC's as the corsairs so.. wierd!!
one other side note is that I have since disabled SLI to try and narrow down the issue.. I will renable it tonight and see if I start to get stability issues again with the mem...
my specs are:
680i P24
2x1GB 8500C5D
2 x asus 8800GTX SLI
silverstone 850w SLI PS
1 x raptor 36GB
X-Fi fatality pro gamer
When i first noticed the issues I would get the C1 error and then have to jockey memory slots configs to geet it to boot, then I went to one mem stick and would get it to boot and set all timings and voltages correctly for the ram.
i have tried all configs already suggested with no luck. i am currently awaiting an RMA board from eVGA.
Since I have bought some crucial 6400 ram and have been running both sticks with out a problem at all...
I dont think that it is the ram, I think that it is the chipset or mem controller...
it doesn't like the corsair for some reason?
although the IC's on my crucial are the same IC's as the corsairs so.. wierd!!
one other side note is that I have since disabled SLI to try and narrow down the issue.. I will renable it tonight and see if I start to get stability issues again with the mem...
Tornadojoe
n00b
- Joined
- Jan 18, 2007
- Messages
- 9
Not to worry,Corsair mem is the best in the business,If the board is doing it,Its going to happen with any mem,I would rather have Corsair with there fine support then brand X,My system is running fine with 680i board and the 8500,I did get the fan kit because it looks so cool and it does work.Working on a rev bios or rev board ? What they going to replace all are boards that are overvolting ?
How long has it been working? I hear it stops after about 90 days of use...
I don't know if you guys visit evga forum but here http://evga.com/community/messageboard/topic.asp?TOPIC_ID=26392 is a whole bunch of people that are having the same issues.
Here's my build:
EVGA 680i
E6300 (stock freq) + Arctic silver+Zallman 9700
EVGA 8800GTX (stock freq)
4GB Corsair 8500C5 initially @ 5-5-5-15@[email protected]
(now only 2GB @800@[email protected] and still unstable)
2xWD 74gb 10K in raid-0, 2xhitachi 40gb in raid-1, 2xwd 140gb in raid-0
Lian-Li Case
PCPower&cooling 750W
Dell 24'' + Samsung 20''
and my problem started one morning (about 2 weeks after perfect operation including hours of 3d on one monitor and movies on the other) when i woke up, typed in my password, opened a webpage and got served with a freeze followed by BSOD. Now, i don't think i ever got a nv4disp BSOD, mine included win32k.sys, page_fault_in_non_page_area, or just all kinds of memory codes 0x0nnnn. It was not always either that I would get a BSOD - like Tigertank wrote - mouse would freeze, then it would look like it would be coming back, sometimes sound would loop, sometimes one of my monitors would become unresponsive while I could still do "right-click" on the other. I mean it was the variety of behaviors that totally confused me. Absence of consistency if you will, or at least the consistency you'd be looking for in system crashes. Oh, I never overclocked anything.
First I blamed OS ( xp pro x64) so to the EventViewer I went. A few errors including pfc.sys and windows search service. Desktop search was uninstalled and pfc removed from LowerFiles with regedit (btw, it took hours to figure out what the pfc.sys was). A few reboots and event viewer was clean. I tried to run AmericaArmy - freeze, Wow - freeze, web browsing - freeze. I was going mad. Before going to hardware i went online and read every forum and support group there was about this sorta issue and ended up reinstalling the system a bunch of times, installing xp pro x86 several times, different drivers etc. I always create acronis backup images so i could roll back o a clean system or stable system - that didn't help either.
Now I turned to hardware - First thing memtest popped 3 errors on moving random patterns (the other 5 tests produced no errors). I pulled every stick out and retested in very slot - one dead. Into the box it went with an RMA sticker. 2 others put back and system ran stable - I was happy. That lasted about a day or two. First it froze on a webpage with some animation on it. Then on WoW, then on IE, then on media player. I ran memtes, goldram and ramcheck, and pc check and nv stress test (only ram/cpu/disk) - all good. I removed the data raids, reinstalled system on a single drive - nothing helped. I was pulling my hair out.
EVGA forum put some hope in me after seeing that others were having similar issues. But whereas some people get stable rigs at lower voltage I still get freezes on 3d and 2d apps. I lowered voltage to 1.85, disabled sli-ready and the system does stay stable longer, but still crashes.
It is good to know that more people are now aware of this and hopefully something can and will be done. I'll be talking to EVGA on monday to rma my board and possibly going to have to rma my other 2gb of ram to corsair.
Here's my build:
EVGA 680i
E6300 (stock freq) + Arctic silver+Zallman 9700
EVGA 8800GTX (stock freq)
4GB Corsair 8500C5 initially @ 5-5-5-15@[email protected]
(now only 2GB @800@[email protected] and still unstable)
2xWD 74gb 10K in raid-0, 2xhitachi 40gb in raid-1, 2xwd 140gb in raid-0
Lian-Li Case
PCPower&cooling 750W
Dell 24'' + Samsung 20''
and my problem started one morning (about 2 weeks after perfect operation including hours of 3d on one monitor and movies on the other) when i woke up, typed in my password, opened a webpage and got served with a freeze followed by BSOD. Now, i don't think i ever got a nv4disp BSOD, mine included win32k.sys, page_fault_in_non_page_area, or just all kinds of memory codes 0x0nnnn. It was not always either that I would get a BSOD - like Tigertank wrote - mouse would freeze, then it would look like it would be coming back, sometimes sound would loop, sometimes one of my monitors would become unresponsive while I could still do "right-click" on the other. I mean it was the variety of behaviors that totally confused me. Absence of consistency if you will, or at least the consistency you'd be looking for in system crashes. Oh, I never overclocked anything.
First I blamed OS ( xp pro x64) so to the EventViewer I went. A few errors including pfc.sys and windows search service. Desktop search was uninstalled and pfc removed from LowerFiles with regedit (btw, it took hours to figure out what the pfc.sys was). A few reboots and event viewer was clean. I tried to run AmericaArmy - freeze, Wow - freeze, web browsing - freeze. I was going mad. Before going to hardware i went online and read every forum and support group there was about this sorta issue and ended up reinstalling the system a bunch of times, installing xp pro x86 several times, different drivers etc. I always create acronis backup images so i could roll back o a clean system or stable system - that didn't help either.
Now I turned to hardware - First thing memtest popped 3 errors on moving random patterns (the other 5 tests produced no errors). I pulled every stick out and retested in very slot - one dead. Into the box it went with an RMA sticker. 2 others put back and system ran stable - I was happy. That lasted about a day or two. First it froze on a webpage with some animation on it. Then on WoW, then on IE, then on media player. I ran memtes, goldram and ramcheck, and pc check and nv stress test (only ram/cpu/disk) - all good. I removed the data raids, reinstalled system on a single drive - nothing helped. I was pulling my hair out.
EVGA forum put some hope in me after seeing that others were having similar issues. But whereas some people get stable rigs at lower voltage I still get freezes on 3d and 2d apps. I lowered voltage to 1.85, disabled sli-ready and the system does stay stable longer, but still crashes.
It is good to know that more people are now aware of this and hopefully something can and will be done. I'll be talking to EVGA on monday to rma my board and possibly going to have to rma my other 2gb of ram to corsair.
Redbeard
Official Corsair Rep.
- Joined
- Nov 3, 2005
- Messages
- 1,859
We are talking to eVGA on this issue right now, trying to get it resolved. They're trying to track it down, but what makes it difficult is that it's a very rare issue, affecting a very small percentage of the people who are using their board. Not only that, it apparently happens with memory besides Corsair, if a lot of other forums are to be believed. Nonetheless, we're going to try and shoot this thing dead before it grows horns and gores some people in the community.
Redbeard you guys have all the bells and whistles, someone just did an interesting test on the evga board - using one stick in blue/black slots trying to get a stable boot. He gets no boot if it is <2.1v in black and <1.9v in blue (here;s the post http://evga.com/community/messageboard/topic.asp?TOPIC_ID=26656). now, i am able to run at 1.85 in either so maybe thats irrelevant but lotsa ppl think its unstable voltage.
I don't have the sales # for EVGA so wouldn't be able to say what constitutes small percentage, but it seems to be growing as all the problems seem to turn up after some time of smooth operation. And EVGA is not the only one - other boards with 680 chipset are also geetting some heat.
What is true - is that Corsair is no the only manufacturer that ppl are having problems with. Pretty much anything running at 2.2+ from what i've gathered. There are suggestions that Micron D9 is somehow related to the issue, but again without knowing what chips are on the stick it is hard to say. I even started a thread for that on evga, but there's not enough data to make conclusions.
All i can say is let's try to keep everyone posted on the issue.
I don't have the sales # for EVGA so wouldn't be able to say what constitutes small percentage, but it seems to be growing as all the problems seem to turn up after some time of smooth operation. And EVGA is not the only one - other boards with 680 chipset are also geetting some heat.
What is true - is that Corsair is no the only manufacturer that ppl are having problems with. Pretty much anything running at 2.2+ from what i've gathered. There are suggestions that Micron D9 is somehow related to the issue, but again without knowing what chips are on the stick it is hard to say. I even started a thread for that on evga, but there's not enough data to make conclusions.
All i can say is let's try to keep everyone posted on the issue.
Go for the headshot Redbeard!
And if it's still kicking after, whip out that nasty Corsair cutlass.
I went ahead and ordered a bad axe 2 to hold me over until evga can figure this out. May as well use it as rebuild piece in this amd system once all this blows over. I still can't get my 8500's to post in the 680i, so they may be dead. I don't know. One stick posts C1 under any circumstance, and the other posts 1D. I've got some low end ddr2 coming with the board this week so we'll see. If I didn't have $250 into it, I'd walk away from 680i completely. Sometimes stability is worth more than having the fastest piece of hardware.
Redbeard
Official Corsair Rep.
- Joined
- Nov 3, 2005
- Messages
- 1,859
It's really not. The issue appears to be confined to long-term usage in the eVGA board at high voltage with memory based on Micron ICs. By long-term I mean 30 days or longer, and by high voltage, I mean 2.2V or higher.
This does not affect all boards, or all memory, in fact my boss's home system has been an eVGA board with Micron D9s @ 2.4V since the board came out, over 2 months ago, and shows no problems at all. I wonder if it's a blue slot/black slot issue. Anyway, we're further investigating, but a friend of mine who works for a competitor of Corsair's has advised me they're seeing the same thing happen on some Micron-based parts on that board.
As soon as I have a guaranteed fix or know exactly what's causing the problem, I'll let you guys know.
AuroraProject
Gawd
- Joined
- Dec 13, 2004
- Messages
- 820
One of my Dominator 1250's gave up tonight, 397 errors in memtest. The other stick is fine. EVGA 680i mobo, 1 week old build running 5-5-5-18 2T @ 2.4v. I just started the RMA process. 
Redbeard, I plan on sending both modules back, but I can hang onto the fan right? And what's the average turn around on an RMA? Do you cross ship? I'm in northern California, so shipping wont take long.
Redbeard, I plan on sending both modules back, but I can hang onto the fan right? And what's the average turn around on an RMA? Do you cross ship? I'm in northern California, so shipping wont take long.
Well, I talked to EVGA today to rma my board and asked some other questions. According to the tech support person i talked to:
1) there is no REV2 of this board.
2) no bios info is available.
3) the capacitors that are missing on the backside were there for some "radiation testing or something" and are removed now.
i'm rma-ing mine, but it is sorta pointless
1) there is no REV2 of this board.
2) no bios info is available.
3) the capacitors that are missing on the backside were there for some "radiation testing or something" and are removed now.
i'm rma-ing mine, but it is sorta pointless
AuroraProject
Gawd
- Joined
- Dec 13, 2004
- Messages
- 820
He's definitely mistaken on the revision issue, I have a revision 2 board. (it doesn't work right either!)
Redbeard
Official Corsair Rep.
- Joined
- Nov 3, 2005
- Messages
- 1,859
One of my Dominator 1250's gave up tonight, 397 errors in memtest. The other stick is fine. EVGA 680i mobo, 1 week old build running 5-5-5-18 2T @ 2.4v. I just started the RMA process.
Redbeard, I plan on sending both modules back, but I can hang onto the fan right? And what's the average turn around on an RMA? Do you cross ship? I'm in northern California, so shipping wont take long.
Not sure on average turnaround on RMA. I really have no ability to peer into the RMA process to see how long it's taking. They will occasionally cross-ship if you ask them very, very nicely though.
AuroraProject
Gawd
- Joined
- Dec 13, 2004
- Messages
- 820
That's probably what it is, I notice many got hard cardboard boxes with the mobo in antistatic bags, mine came in a flimsy paperboard box and a formed plastic insert with the mobo inside.
Redbeard
Official Corsair Rep.
- Joined
- Nov 3, 2005
- Messages
- 1,859
Because of the reported nature of the problem (ie - failure after long periods of use), it's going to take a while to test for it. Long-term testing takes time. We haven't seen any of our newer lab systems fail yet, but we're working with both eVGA and nVidia to determine what exactly is happening.
It's in all our best interests to get any problem that might exist identified and then create a solution as fast as possible. We all want you guys to be happy with your purchases and have the best motherboards and memory available on the market.
We're going to have some semi-long-term tests going for the next week or so, and I'll get you guys some results as soon as we have something valid and repeatable.
I was just wondering what you suggest we do at this point. Since the problem has been acknowledged, should we just wait? I'd prefer to deal with an RMA through my retailer which only lasts for 30 days (I may even be slightly out of that period now); however, I don't want to just RMA the motherboard only to run into the same problem again and waste everyone's time.
Thanks Redbeard.
Thanks Redbeard.