Jump to content
Corsair Community

Gather 'round, folks with TWIN2X4096-8500C5D problems


YCH

Recommended Posts

After noticing increasing BSOD occurances over the past month, I began suspecting the motherboard/CPU/RAM. Then I finally decided to run a series of methodical tests over a span of two weeks to ascertain the component at fault.

 

I'm too tired to list all the tests right now, but the most telling one is that setting my system memory speed to a pretty standard 800MHz, 2.1V and 5-5-5-15 latencies and keeping them unchanged througout, my old set of TWIN2X2048-6400C4D runs just fine without random BSODs or memtest errors, while the TWIN2X4096-8500C5D flakes out seemingly at random. However the symptoms are too similar to those reported by various people on this forum to be pure coincidence.

 

Basically, here is the pattern I've found with folks that have problems with the TWIN2X4096-8500C5D (including myself):

 

- System runs stably for the first couple months. Memtest and other stress tests do not show errors.

- System starts to randomly BSOD during near idle moments. Eg: leaving the computer overnight etc and come back to see a blue screen.

- Trying to reproduce the conditions for errors is an exercise in frustration and futility. Memtest would run for hours on end without problems. Then you continue to use the PC normally thinking you've found a stable setup...then while near idle the PC BSODs on you. You immediately run memtest and find a bajillion errors. You turn off the PC and back on again and the errors may be gone and memtest runs stably again. Before a BSOD strikes yet again later.

 

Alright folks, tell me this doesn't sound familiar.

 

I'd like to think of myself as a competent person at troubleshooting, and I always isolate variables and test them one by one if possible, resisting the urge to speed through and hope for a silver bullet. Yet the nature of this problem is very elusive. There are far too many BIOS settings and variables for me to feasibly run through, and many of them insufficiently explained. I don't have an oscilloscope at home so I have no chance in hell to look at edge skew, voltage ripple, line reflections and all that.

 

But surely good RAM shouldn't require aligning the planets to find that stability at pedestrian speeds? Again, I'd like to reiterate that by doing nothing more than powering off the PC, taking the TWIN2X4096-8500C5D and replacing it with my older, slower TWIN2X2048-6400C4D system stability is much much higher. I think at the very least this points to some kind of incompatibility between my Asus P5K Deluxe and the TWIN2X4096-8500C5D. And if other people's reports are true, then this RAM with many other motherboards as well.

 

Now I intend to RMA these units, but I am loathe to actually use them when I receive a new pair because many folks here seem to have the same problem even on their 2nd or 3rd set.

 

The big problem is that the instabilities are hard to reproduce yet the general pattern/signature of the failure is recognizable and potentially dangerous. I think I've had data corruption on one of my OS'es now due to the repeated BSODs I've tried to force out of the system over the past two weeks of testing.

 

So I pose this question to you Corsair support staff:

 

Are you guys investigating this issue to understand the problem and possibly pinpoint the root cause? It's not just to appease customers but it'll sure come in handy for future products so whatever this mysterious thing is doesn't plague Corsair again.

 

Finally, here are a few links to other people having similar issues:

http://forum.corsair.com/forums/showthread.php?t=83350

http://forum.corsair.com/forums/showthread.php?t=83312

http://forum.corsair.com/forums/showthread.php?t=81545

http://forum.corsair.com/forums/showthread.php?t=82646

http://forum.corsair.com/forums/showthread.php?t=82810

http://forum.corsair.com/forums/showthread.php?t=82247

 

edit: I have found a pretty reliable way (for myself at least) to test for this specific problem. I'd run a data backup (using the simple xcopy command in windows) from one of my internal harddrives to an external USB harddrive overnight...and usually this is enough to provoke a BSOD from an unstable RAM/motherboard combination.

Link to comment
Share on other sites

  • Replies 83
  • Created
  • Last Reply

More excerpts of some negative feedback on this set of RAM on Newegg.com (they all seem to say roughly the same thing???). And this is for feedback within the last couple months. I didn't look further.

 

Pros: Cool heatsink on DIMMs, OK WHILE they worked.

 

Cons: 7 months in, dead as doornails. Decent support, but who has the time? You spend the $$, install and verify. S/B end of customer problem, then they fail, and you're asked to do back fkips for 40 hrs to solve their QC issues. Bad Corsair, Bad!

 

Other Thoughts: Vista 64 bit(yeah, I know...), GA-EP45-UD3P, 8 GB of this stuff (4 sticks), Corsair 650w PSU. Corasair support site rife with this complaint.. 6 months in belly up, Sounds like many RMAs worse than 1st complaint. All sticks fail memtest now.

 

Now I gotta eat and buy new memory that works.

 

Pros: -Was fast when it worked.

 

Cons: -Stopped working after a few months.

 

Other Thoughts: This is mainly a complaint to Corsair. I RMAd my original set because it became defective. I received a defective kit and sent that back as well. I just received another kit which also has both stick defective. So 3 out of 3 have been defective. I've had to pay an extra $20 in shipping charges. This is what drives customers away. Sending bad product in exchange for bad product. I work in retail and if I were a customer at my store, I highly doubt I'd be going back to this kind of company,

 

Pros: Looks nice. Thats about it.

 

Cons: This memory stopped working and caused BSOD's then a constant reboot cycle after memory dumps after only two months. This was at stock timings, stock recommended voltages. Never overclocked. And my case is very well ventilated, so I know it wasn't a heat issue (I have a Lian-Li PC65B with (5) 80mm and (2) 120mm fans...Quad Core processor which runs at mid 30's C at idle). On another but similar note, I also have a 4Gig Corsair USB memory stick that is a little over a year old that suddenly dumped all the info that was on it. I haven't bothered to see if it still works, I am going to RMA both the USB stick and these desktop memory modules.

 

Other Thoughts: Corsair used to be a great company when it came to memory. My experiences with this company's products over the last year have been horrible. I lost alot of important information I used as an instructor because of the failed memory stick. I am only thankful that the constant reboots due to the bad desktop memory didn't corrupt my O.S. or hurt any other components such as my motherboard or hard drives. I will never purchase another memory product from this manufacturer again, and I cannot recommend to anyone else to do so either.

 

Pros: none

 

Cons: For the high price, this should work. This memory had countless errors in Prime and with Memtest. It wouldn't work in dual channel mode either. I bought a set less than a year ago. It started to have numerous errors. This was a replacement set and it had nurmerous errors "out to of the box". I just returned it.

 

Other Thoughts: Newegg.com gave me a full refund and even called to see to make sure I was happy. A great company.

 

Motherboard was a ASUS P5Q that is now working great with [ZOMG_EDITED] memory.

 

Pros: Ok new set has a stick so bad it keeps my computer from booting. At least the first set would work most of the time.

So after I ship the first set back, I will then ship the new set back. At this fail rate I will no doubt have to ship at least one of the replacements.

 

Cons: How many sticks can a ramship ship if a ramship could ship ship. Oh brother, so I,m looking at $20.00 in shipping fees.

 

Other Thoughts: About the rebate debit card. First always mail it with return reciept request, that way they can,t say they didn,t get it.

Then try and find a place that will actually take it.

 

What,s your time worth this memory is good when it works, but.

Link to comment
Share on other sites

Keep in mind that you are seeing results of one of our most popular memory kits over the past 2+ years. There are literally tens of thousands of these kits out in the world. The fact that you have found a small number of issues in a support forum is logical. People come here when they have issues. There is no underlying defect here.
Link to comment
Share on other sites

Keep in mind that you are seeing results of one of our most popular memory kits over the past 2+ years. There are literally tens of thousands of these kits out in the world. The fact that you have found a small number of issues in a support forum is logical. People come here when they have issues. There is no underlying defect here.

 

How can you say that with such certainty? Can you back that up with statistics - indulge me with some statistical analysis (eg % return due to this failure vs all returns of this item... % returns of this item vs other RAM kits of similar sales volume and customer base). Please don't make such strong statements unless you have something to show for it.

 

I'm usually pretty savvy when buying computer hardware. I know when to take into account one off type failures and not have that affect my purchase decision. However, what I do look for is failure trend with specific details (not comments like "item X does not work" but instead "item X fails in Y fashion under Z conditions"); ie if more than a few people within the last few months report a similar type of failure I'll stay away because it indicates a likely design issue with a product. This time, I made an exception because it was Corsair (yeah that's the level of faith I had - been a Corsair customer for years and years) and I decided it was worth the risk at $85 back then.

 

Turned out it was more trouble than it was worth.

 

BTW, shrugging off an uninvestigated issue is bad engineering practice in general. Without ruling out design or manufacturing problems, you are just opening yourself up to possible backlash in the future.

Link to comment
Share on other sites

How can you say that with such certainty?
Because I see the actual RMA figures. I see our support forums, our internal tech support statistics, and I surf 25+ enthusiast forums daily. We've been making these modules for a long time and there are a lot of them out in the world. The fact that you see issues in a support forum, as I already said, is logical.

 

A certain percentage of any electronic device will fail, no matter how much testing you do. We monitor all of these factors. EVERY XMS module is hand tested in a motherboard to run its rated spex before packaging.

 

There is no underlying defect here.

Link to comment
Share on other sites

Because I see the actual RMA figures. I see our support forums, our internal tech support statistics, and I surf 25+ enthusiast forums daily. We've been making these modules for a long time and there are a lot of them out in the world. The fact that you see issues in a support forum, as I already said, is logical.

 

A certain percentage of any electronic device will fail, no matter how much testing you do. We monitor all of these factors. EVERY XMS module is hand tested in a motherboard to run its rated spex before packaging.

 

There is no underlying defect here.

 

Thanks, but you're not really telling me anything I don't already know.

 

The only way you're going to convince me is to show me the RMA data distribution, which is obviously not possible since that is confidential data.

 

Also, just because you test the RAM kit to run on some motherboards to run its rated specs doesn't mean you've caught all possible issues. I'm just asking Corsair to look into this frustrating kind of failure. I'll accept device failure under high stress conditions (eg overclocked or non documented use). But not when the device fails under low stress way below specified ratings, documented by numerous users.

Link to comment
Share on other sites

Memory modules, and any other device using semiconductors, do not need to be overclocked, stressed, or run beyond specification to fail. All they need is current. If what you are saying were true, no OEM computer system, no television etc etc etc would ever fail. But, they do. And, when ours do, we replace them.

 

We appreciate customer concerns but there is no issue here. We have been in the memory business for a LONG time now and we know how to manage issues and quality control.

Link to comment
Share on other sites

We appreciate customer concerns but there is no issue here. We have been in the memory business for a LONG time now and we know how to manage issues and quality control.

 

I'm sorry to disagree Mr Yellowbeard, these RAM modules have caused nothing but problems, has costs me a lot of money and wasted my time and not mention the hundreds of users out there that are experiencing the same issues. I've been searching the net to try and find results and I just continue to find people nearly in tears because they don't work. From what I am reading it sounds like there is a possible engineering fault in the IC chips that are being used. The quality of Corsair products is great but if you collate all the information that end users are experiencing, you will notice that there is a pattern appearing. It's the same story over and over and over again and it should be raising alarm bells.

 

Something could very well be wrong with this particular IC chipset and it needs to be escalated to the engineering level rather than left for a customer service officer to deal with.

 

You state that you see the stats in regards to RMA's, I'm curious whether the the RAM Modules being returned are of the same IC chip?

Link to comment
Share on other sites

I also de think there is an issue here. I've been traveling from the msi (my mobo) to the AMD (my procc.) to this forum for this problem to be solved. On all forums the same problems with this type of memory. A lot of people think they have it solved and then after a week or 2 the bsod start again. I bought myself a new harddisk because i really thought i had found it now. but is wasn't the harddisk. I have spoken with someone who thought he had found the problem by replacing his powersupply. but again same problems came back after a few weeks.

 

I also think that not all people who use the memory and experiance problems come to this forum to post. maby they just read it and decide to buy other memory. I think it is realy hard to get good statistics.

 

I hope Corsair comes with a sollution.

 

Regards,

Alex

Link to comment
Share on other sites

  • Corsair Employee
There is nothing we can change on the modules other than offering you a replacement with an RMA but if that does not solve the issues then there is likely some other problem. And I would suggest talking to the MB maker to see what they have to say. But WAPC this part is one of the most popular parts we have made and we have sold Millions of sets so as Specmike told you there are bound to be users who who have common problems that is fairly common with large volume parts and the return rate of that part has been about 2% +- 1.0% throughout the production cycle knowing how sensitive some MB's and CPU's have became would be a more likely cause of similar failures.
Link to comment
Share on other sites

Count me in as another casualty to the 8500C5D failure...:[pouts:

 

Running fine for 2 months from new with an ASUS M3N-HT Deluxe, now the same very frustrating symptoms as already reported by many others.

 

As the issue has appeared across a number of different brand motherboards (and I now know of one case where different branded ram has solved the issue) I think it's safe to figure this particular RAM does have an issue. That the fault recurs with replacement ram clearly suggests we are not looking at a 'bad batch' either, but an inherent problem (after all, how likely is it that all the different motherboards involved could all cause the exact same problem?).

 

Looking at the RMA numbers posted above I can understand why the official line is to suggest such an inherent issue is not the case, but when the same problem occurs across different system configurations I do not think Corsair can ignore the possibility they have dropped the ball to a certain extent with this RAM.

 

Though I have already put a RMA in process for my own RAM I think I am going to have to look at something else.

Link to comment
Share on other sites

I agree with the others that this RAM has definitly a problem that has to be adressed. Got a pair one month ago from an Online store and had to ask for a replacement 2 days after installing it in my system. Got the new modules and had to ask for a second RMA because one of the modules was immediatly showing errors. I am actually waiting for this second RMA and i honestly doubt that it will finally work. Last chance for COrsair. If this one does not work, i will consider this hardware as being a loss. I will stop paying for shipping fees that will ultimatly end to nowhere. Honestly guys , it is one thing to say that there are thousand of these out there, but look at the percentage of dissatisfied customers and you will see that you are below average on that one compared to your competition. I am not a newbie in system building, i have been doing this for 20 years and honestly i have never seen that from a reputated manufacturer whose product is on the QVL of my motherboard.
Link to comment
Share on other sites

It seems to me you guys want something thats 100% perfect. I to had to send in bad 8500c5d but new ones work good i use this on 2 other system builds and no problems

 

I want something to last more than a couple of months before failing and screwing up my system. That this particular problem* clearly does happen again with replacements leaves us facing the prospect of having to RMA the ram every few months. Can't say I find that in any way acceptable.

 

*btw, were you affected by this problem?

Link to comment
Share on other sites

yes i had 1 bad stick in a 4 gig kit got my new ones and now everything seems good my other 2 builds don't have any issues im using older gigabyte mother boards. The mother board i had the biggest issue with was 780sli mother board did not take the memory very well. no matter what it didnt run at the specs it said it was couldn't get it past 900 no matter what i did but the 3 gigabyte boards im using work just fine with pc8500C5D so im staying with gigabyte.
Link to comment
Share on other sites

There is nothing we can change on the modules other than offering you a replacement with an RMA but if that does not solve the issues then there is likely some other problem. And I would suggest talking to the MB maker to see what they have to say. But WAPC this part is one of the most popular parts we have made and we have sold Millions of sets so as Specmike told you there are bound to be users who who have common problems that is fairly common with large volume parts and the return rate of that part has been about 2% +- 1.0% throughout the production cycle knowing how sensitive some MB's and CPU's have became would be a more likely cause of similar failures.

 

Hey RAM GUY,

 

I work in the semiconductor industry, and I know how tempting it is to sweep something that is within the "acceptable range of failure rate" under the rug. Simply because it's easier and cheaper in the short run. I also know that failure information given along with an RMA is...how should I put it...not entirely dependable...unless the customer has run extensive cross-checks and tests prior to the RMA. Not sure how it is with retail products like this but I'm guessing the information you receive on the whole is not as good as you would like...which means telling how much skew or trend in failures of a product line is challenging. PS: I realize how much of a know-it-all I sound like at this point but please hear me out; I can only report from what know.

 

So despite your assurances of a low failure rate of this particular part, I still urge you guys to have a real look at issues like this, precisely because you cannot explain to us what the failure mechanism is in this case (while your customers are all telling you they have a particular pattern in their failures). Honestly, if this problem is identifiable on the day of purchase via stress testing, I would have a lot less problem with it because I would know off the bat the RAM is good or bad and have it replaced immediately if is bad. But it looks like that the failure we're talking about is intermittent, and not obvious until a month or two into the use time.

 

It just might be worth your time to spend some R&D indentifying the problem and working with motherboard makers (we're talking MAJOR players here like Asus, Gigabyte etc so it's not some small fry you want to have problems with). As retail customers, it's not our responsibility to do in-depth QA and testing for you to identify problems. In fact, you should be thanking us for bringing what we feel is a serious problem to your attention - and take steps to investigate. Imagine what it would be like if customers didn't bother returning defective RAM and didn't bother telling you what's wrong with it. You'd be blind if something goes wrong.

 

I love Corsair, but that love only stays as long as their products deserve it.

Link to comment
Share on other sites

I think at the very least this points to some kind of incompatibility between my Asus P5K Deluxe and the TWIN2X4096-8500C5D. And if other people's reports are true, then this RAM with many other motherboards as well.

 

Out of curiosity, what CPU are you using?

Link to comment
Share on other sites

When you look around at each case that has this same particular problem and can see the only common factor is this model of ram it does seem clear there must an issue with this model and certain motherboard chipsets. Not everyone is getting this, and I would guess there are many of them with no problems who have the same chipsets as those who are afflicted as well, so isolating the exact cause at our end is pretty much impossible.

 

When we also know that once a system develops this problem it does not go away without replacing the ram we also know something is happening that is causing a permanent physical change/damage to the ram as well. It may be only be a subtle change as cold boots will correct the issue temporarily, but it is always there. Knowing as well that this issue will not occur again when using different ram suggests the remaining hardware is not where the problem lies at all.

 

Just what change/damage this is I have no idea, but I do know after 2 months of solid stability my system at the moment is a constant time bomb, and I regularly shut it down and pull the power (every couple of hours) to try and avoid another crash.

Link to comment
Share on other sites

hi!

 

i have exactly the same problem with my 2 month old ram and i have done a lot of testing to narrow down the problem, but i am at the end of my knowledge right now.

 

for me the problem always starts with a bsod. the longer the computer is in operation the more likely it will happen. after this point there is no way to boot up windows again (bsod) and also memtest shows allot of errors.

but after a full power cycle (off/on) everything is rock stable again until to the next bsod.

 

so i have tested the sticks one by one in slot 1 - 40 passes no problem

and i have tested the sticks in dualchannel - 40 passes 600.000 errors

 

due this, i decided to replace my mother board and my computer was rock stable again (prime95 blend 14h stable, memtest 1000% no error, 2x memtest 86+ 40passes 16h no errors, linx 30x no error).

 

after 3 days of happiness - bam bsod - rock stable to total system failure.

 

edit:

asus rampage formula (replaced)

TWIN2X4096-8500C5

intel q9550

corsair hx650

Link to comment
Share on other sites

Changing the motherboard and getting this issue again is more confirmation that the Ram itself does develop a permanent fault from this particular failure.

 

I just hope the motherboard itself doesn't get compromised as well by the fault. That your new board was ok for 3 days before crashing might suggest our boards are not escaping future issues from this ram's inherent problem.

Link to comment
Share on other sites

i don't think a faulty cpu will pass 100 times linpack, which is the same program used by intel to test their own cpus (recommend are 5-10 and not more than 20, because of the stress). on the other hand this also proofs that the allocated ram by the program is fine at the moment. but fact is that something is wrong here and the symptoms are bsods followed by a memory failure on a reboot. i mean its like the whole ram is failing. the bsod happens almost at idle (loadline calibration is disabled) and a cold boot will fix it for the next hour, days, who knows.

 

http://members.chello.at/mrorange/linx.jpg

Link to comment
Share on other sites

mrmrk: Sounds like it could be the CPU, but there's no real evidence one way or another.

 

I don't think the CPU is likely to be the problem as they are getting the same particular symptoms as many others with AMD CPU's and different boards.

 

Generally, when you are presented with the same problem happening in different cases you look for something that is present in each of those cases for the cause.

 

From everything I have read the only common factor for those getting this problem is the Ram.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.


×
×
  • Create New...