Corsair 6400C4 fails after 25 minutes

[wdn]

Run at the rated 800 MHz, 4-4-4-12 timings and 2.1V operation, this set of Corsair TWIN2X2048-6400C4 lasted for no more than 25 minutes in Memtest-86 before it started throwing memory errors. Here is a actual screen shot of the monitor I took of it running just a few minutes ago. Not sure how soon it failed, I started Memtest went in the living room to watch a little TV, came back during the commercial break and the screen was already red. Who knows? it could have bombed during the first 5 minutes for all I know.

No surprises there, these sticks never ran stable at more than 350 MHz FSB clock (i.e. 700 MHz) in Windows. The memory is about 6 weeks old. With the PROMOS IC's, of course.

I already have been issued an RMA from newegg. So other than the cost of sending back these junk sticks at my expense and being down for a week or more, I guess I am all set.

Here is the Overview for these memory sticks directly from newegg website. Quote without comment:


"Corsair's XMS (Xtreme Memory Speed) memory has long been popular among enthusiasts. Corsair has many such kits, and all are regarded as some of the best available in their categories. With 2GB (2 x 1 GB) of memory in dual channel configuration running at 800MHZ speed, this CORSAIR XMS 2GB DDR2 SDRAM System Memory, offers targeted low latencies of 5-5-5-18 at JEDEC standard speeds and have been certified as “SLI Ready”, giving you some screaming data transfer rates and exquisite renderings.

These modules are made by Corsair using carefully inspected and matched RAMs. Running in dual channel mode, the module provides 12.8GB/s worth of memory bandwidth which happens to be exactly what the dual core Socket AM2 Athlon64 X2/FX processors need. What a coincidence. This is memory that’ll both overclock and stay cool, thanks to their huge black heatspreaders. This is a great kit for anyone wanting to max out their system’s performance!"

 

[Redbeard]

Bad memory

I'm a bit confused. Haven't we already seen your comments about your bad memory here and here? You got bad memory, it happens. If you hadn't already started an RMA with Newegg, you could easily RMA them through us.

I'm confused as to what your complaint is right now. You got bad memory, that sucks, I'm sorry. It's getting replaced. Is there something I might be missing?

 

[SentToSchool]

I'm a bit confused. Haven't we already seen your comments about your bad memory here and here? You got bad memory, it happens. If you hadn't already started an RMA with Newegg, you could easily RMA them through us.

I'm confused as to what your complaint is right now. You got bad memory, that sucks, I'm sorry. It's getting replaced. Is there something I might be missing?

He's angry because he has to pay shipping back and wait for many days to get a replacement.

 

[Damuman]

I got reimbursed for shipping =D

 

[wdn]

I'm a bit confused. Haven't we already seen your comments about your bad memory here and here? You got bad memory, it happens. If you hadn't already started an RMA with Newegg, you could easily RMA them through us.

I'm confused as to what your complaint is right now. You got bad memory, that sucks, I'm sorry. It's getting replaced. Is there something I might be missing?

1. Paying return shipping on your defective product. Your mistake, my dime. This was not the luck of the draw this was a bad batch of sticks and a bad QA process.

2. Down for at least 7 days because of said defect. What you don't cross-ship even after a fiasco like this?

3. "Blame the customer" "Marginalize the customer" Re-read some of your earlier responses.

Our RMA rate is extremely low, since XMS is tested for functionality before shipping. We don't know what's going on yet at Newegg, whether it's a legitimate series of bad modules, a manufacturing defect, a bad batch of ICs, or maybe it's just somebody having a lot of fun with the newegg reviews of our products.

DOES THIS LOOK LIKE I'M HAVING FUN ?? More fun than a barrel of monkeys isn't it, hmm? Since you seem to be questioning the integrity of newegg customers and their reviews let's set the record straight shall we?

Let's just say after a swipe like that against the customer's integrity it is well worth my time to drag out the digital camera and set the record straight for everyone to see the truth for themselves.

We don't know, and that's why we asked Newegg to ship those parts back and allow us to replace them with parts we'd tested already. The parts you buy from Newegg now have been fully tested.

OK I believe what Corsair says now, I suppose my mistake was in also believing Corsair then all that stuff about "carefully inspected and matched RAMs" and whatnot and the premium price to go with it.

 

[Damuman]

1. Paying return shipping on your defective product. Your mistake, my dime. This was not the luck of the draw this was a bad batch of sticks and a bad QA process.

Yeah, they really need to stop acting like these are isolated incidents, and that we're being unreasonable because we're complaining. and that they're doing us a favor by rma'ing it.

 

[Redbeard]

Yeah, they really need to stop acting like these are isolated incidents, and that we're being unreasonable because we're complaining. and that they're doing us a favor by rma'ing it.

I believe you guys have bad parts, and I'm sorry if I come across like I'm marginalizing you guys. That's not my intention. I know there are bad parts out there, anybody who's ever sold anything has sold bad or faulty parts. It's going to happen.

We do offer cross-shipping occasionally if the customer requests it. We've done it before and can do it fairly easily.

My point with the newegg reviews is this: we don't have a higher RMA percentage on the 6400C4 than we do on any other part. Also, the 6400C4 is not having the same bad reviews at other sites, like ZipZoomFly. This leads me to question the validity of all the bad reviews being placed on Newegg. I'm not questioning you guys, you seem to know what you are doing and testing it in a matter-of-fact manner. But when you see 6 bad reviews in 6 months, and then within another month you see 12 bad reviews, you start wondering what is happening.

If our RMA rate had gone up as well, we'd be more concerned, but the same percentage of people are RMAing the part as are RMAing our other parts, within the margin of error. Our PROMOS-based parts do not have a higher RMA rate than our Micron-based parts.

 

[Damuman]

I believe you guys have bad parts, and I'm sorry if I come across like I'm marginalizing you guys. That's not my intention. I know there are bad parts out there, anybody who's ever sold anything has sold bad or faulty parts. It's going to happen.

We do offer cross-shipping occasionally if the customer requests it. We've done it before and can do it fairly easily.

My point with the newegg reviews is this: we don't have a higher RMA percentage on the 6400C4 than we do on any other part. Also, the 6400C4 is not having the same bad reviews at other sites, like ZipZoomFly. This leads me to question the validity of all the bad reviews being placed on Newegg. I'm not questioning you guys, you seem to know what you are doing and testing it in a matter-of-fact manner. But when you see 6 bad reviews in 6 months, and then within another month you see 12 bad reviews, you start wondering what is happening.

If our RMA rate had gone up as well, we'd be more concerned, but the same percentage of people are RMAing the part as are RMAing our other parts, within the margin of error. Our PROMOS-based parts do not have a higher RMA rate than our Micron-based parts.

Well the corsair reps do seem like they're telling the truth, so this is what I think is happenning.

I think a bad batch of memory got sent to newegg, and that ours is not an isolated incident.

However, I do believe what you're saying is true about the ratio of purchased ram to rma'd ram, since this is probably THE most bought ram for all these new c2d systems. It's possible the problem isn't as big as it seems, because people who have the ram working fine are less likely to say anything about it, whereas people who don't have it working correctly are most assuredly going to say something.

I had no intention of overclocking my ram in the first place, I'm perfectly happy to sit at 401 fsb with my c2d at 3.2 ghz. However, I understand why the people who bought the ram with the intention of some serious overclocking are mad. XMS has been promoted as performance ram, and has been so in the past. This and the 6400c4d line are the ones not living up to that name, but there is a legitimate reason.

Since this ram is probably the most in-demand, corsair could not have possibly gotten enough micron chips to keep up with the demand. So my theory is that corsair decided to use promos chips to keep up with the demand, while offering that $40 rebate that has been up for as long as I can remember, to make up for it.

Now I think at this point, the issue that will directly affect corsairs sale of this ram is:

When the rebate expires, the ram will still be promos chips. There's so much highly overclockable micron d9 ram available at the pre-rebate price (~$280) that it's really illogical to get the corsair 6400c4 and c4d, without the rebate. I seriously doubt they'll change back to micron chips.

Now of course there's always gonna be people who don't know the difference, but is corsair just going to rely on that? Obviously there's not gonna be a shortage of micron d9's forever, so their actions will prove if this indeed was a "bait and switch," as so many people are saying. I'm extremely interested in a response to this, as this would really determine the type of company corsair is.

edit: I also just noticed this. Even the people leaving 4 eggs on newegg had bad sticks, and over a large span of time, so I don't think it was a single shipment that was wrong. I also don't think this is isolated to newegg, or corsair for that matter. The problem appears to be the promos chips in general, because the failure and doa rates on promos ram is so much higher then with every other type of chip, and it is consistent on EVERY promos chip. In contrast EVERY micron based ram has little to no reviews stating anything wrong.

 

[Redbeard]

Now I think at this point, the issue that will directly affect corsairs sale of this ram is:

When the rebate expires, the ram will still be promos chips. There's so much highly overclockable micron d9 ram available at the pre-rebate price (~$280) that it's really illogical to get the corsair 6400c4 and c4d, without the rebate. I seriously doubt they'll change back to micron chips.

There are a lot of unsorted Micron-based parts around that price point, yes. We don't see any value in doing that when our current sorting process has to yield 5 separate high-end Micron-based parts. (6400C3DF, 8500C5D, 8888C4DF, 9136C5D, 10000C5DF).

If other manufacturers were to implement screens and screen high end parts out, they'd be selling their parts like we are. However, there are only a handful of names as strong as Corsair in the memory field, and if you're going to spend $500-600 on a set of memory, you want a name you can trust. Corsair and maybe 3-4 other brands exist in that field. So when a 2nd or 3rd tier manufacturer gets some Microns, their most logical course of action is to build up modules with unsorted ICs and sell them at 667 or 800 MHz JEDEC spec, and just rely on the word-of-mouth and random overclocking results from the community to sell them. And that's fine, it's a viable business model, it's just not ours.

We won't change to Micron chips because we never used Micron chips. I've said this before, the only 6400C4D modules with Micron ICs on them are modules that failed our 8500C5D screen and had to be rebadged at a lower speed. This is about 1% of all 6400C4D parts.

Now of course there's always gonna be people who don't know the difference, but is corsair just going to rely on that? Obviously there's not gonna be a shortage of micron d9's forever, so their actions will prove if this indeed was a "bait and switch," as so many people are saying. I'm extremely interested in a response to this, as this would really determine the type of company corsair is.

There actually might be a bigger shortage than everyone expects. Micron isn't making a whole lot of money in the desktop memory IC business. They are a US-based company and labor and manufacturing costs are high, with profits being relatively low compared to some of their other businesses. I wouldn't be surprised to find Micron pulling out of the market entirely in a year or two if they wanted to. They could easily spend that capital elsewhere and make more profit.

The reason that the D9s are in short supply is because of the demand. They're the only real "high-end" part right now, especially when screened. If PROMOS's die shrink performs well, and the newer Elpida parts perform well, we might see some new parts with new specs selling for different price points. We have no intention of switching the 6400C4D to Micron, we sell tens of thousands of the 6400C4D a month and cannot make that part with Micron ICs, as we just can't get enough of them and the price point wouldn't make sense. Selling a part like that with unsorted Micron doesn't fit into our business model very well, and we're content to let our competitors fill that niche.

The problem appears to be the promos chips in general, because the failure and doa rates on promos ram is so much higher then with every other type of chip, and it is consistent on EVERY promos chip. In contrast EVERY micron based ram has little to no reviews stating anything wrong.

This is factually incorrect. Our internal RMA rates on the 6400C4D (Promos-based) and the 8500C5D (Micron-based) are nearly identical, and aside from those Newegg reviews, none of our other retailers are noticing higher-than-average DOA or RMA.

It's still possible Newegg got a bad batch, which is why we brought them in for testing. We didn't need to do that, but we felt we'd rather be 100% sure. When I find out the results from our testing, I'll tell you guys. If they were bad modules, I'll let you know. If they pass all our internal screens, I'll let you know that, too. It's entirely possible that because of the popularity of the part (it's many times more popular than the slightly higher part, the 8500C5D), and the drama surrounding the Micron/Promos IC scenario people came to equate the 6400C4D with failure, and began reporting more on it on Newegg.

I have no problem believing the guys on this forum, you guys know what you're doing for the most part and I'm here to help and such, but I've seen stuff on Newegg before that was just sour grapes, and I would hope nobody would be that petty.

If you guys have personal questions or want more info, feel free to PM me.