Corsair Dominator 9136 Preview.

[Kunaak]

I have been monitoring newegg's stock of the 8888 part through auto notiy and I check every day. It's been months since I have seen it there. I am not going to buy it because I can't see spending that much on memory, but the fact that one of the largest online retailers of computer hardware and parts can't get it for that long says something to me.

well your right... it should say something to you.

look somewhere else.

over 10,000 of these have been shipped somewhere and regularly sell out....
someones bought them somewhere.

[Dredd]

well your right... it should say something to you.

look somewhere else.

over 10,000 of these have been shipped somewhere and regularly sell out....
someones bought them somewhere.

uh huh...

the number of units shipped does not mean they're all sold. It's just like Sony or microsoft comming up with magical numbers of "console units shipped".

I don't really care one way or the other, but all the big online retailers that have good shipping prices and people know...can't get it. Explain that

[Kunaak]

uh huh...

the number of units shipped does not mean they're all sold. It's just like Sony or microsoft comming up with magical numbers of "console units shipped".

I don't really care one way or the other, but all the big online retailers that have good shipping prices and people know...can't get it. Explain that

if you can't find it somewhere, look somewhere else.
whats so hard to understand about that?

some people have thier dominator ram, where do you think they get it?
some one bought it somewhere.
I dont get whats so hard about that concept...

and your right, shipped units does not mean all sold...

which again, is exactly why I said, if you cant find it somewhere, look somewhere else.

[Redbeard]

Shipped =/= sold

It does when everywhere we shipped them to has sold out of them almost instantly.

I'm not trying to be smarmy, but what do you think is happening?

Do you think we released a part where we can only make like 5 of them? What possible benefit would that be for us? Hype for an unreleased product? Other companies have done this (Duke Nukem Forever, Starcraft: Ghost, the Phantom Game Console) and the community looks at those products as a joke now. Ghost was even cancelled.

I'm sorry you're having difficulty finding them, I really am, but dude, they're selling. They're online at other retailers occasionally, I know this for a fact. Newegg has sold them. ZipZoomFly has sold them. A bunch of European stores have sold them.

If you have problems with an online store not stocking them frequently enough, email the store and tell them to order more of them. I'm on your side

You want to buy our product, I want to sell you our product. We don't sell direct, so the only way to do it is through retailers. If Newegg orders 10 kits and sells out in 5 minutes, how is that our fault?

I'm not saying that's what's happening, but you have to give us the benefit of the doubt here. We're making them, selling them, and shipping them. They're in the hands of people on this very forum that bought them from those sites.

Are they hard to find? Yes. But so are PS3s and Wiis.

[pentium777]

It does when everywhere we shipped them to has sold out of them almost instantly.

I'm not trying to be smarmy, but what do you think is happening?

Do you think we released a part where we can only make like 5 of them? What possible benefit would that be for us? Hype for an unreleased product? Other companies have done this (Duke Nukem Forever, Starcraft: Ghost, the Phantom Game Console) and the community looks at those products as a joke now. Ghost was even cancelled.

I'm sorry you're having difficulty finding them, I really am, but dude, they're selling. They're online at other retailers occasionally, I know this for a fact. Newegg has sold them. ZipZoomFly has sold them. A bunch of European stores have sold them.

If you have problems with an online store not stocking them frequently enough, email the store and tell them to order more of them. I'm on your side

You want to buy our product, I want to sell you our product. We don't sell direct, so the only way to do it is through retailers. If Newegg orders 10 kits and sells out in 5 minutes, how is that our fault?

I'm not saying that's what's happening, but you have to give us the benefit of the doubt here. We're making them, selling them, and shipping them. They're in the hands of people on this very forum that bought them from those sites.

Are they hard to find? Yes. But so are PS3s and Wiis.

ZipZoom had 9136 I think 20 sets of them yesterday, I watched last night and through today, finally around 4pm they were all sold, the 6400C3DF part is really good and as of around 5:30pm they still have stock. In fact there appear to be around 14 units still for sale. From what I've been told 6400C3 is 2nd hardest to make only to the 8888C4. From third party sites the older 6400C3 (before dominator DHX) would clock up to 1100 4-4-4-12 and 1130 5-5-5-15 but with DHX addition I would expect them to do even better plus the fan is included.

[Dredd]

It does when everywhere we shipped them to has sold out of them almost instantly.

I'm not trying to be smarmy, but what do you think is happening?

Do you think we released a part where we can only make like 5 of them? What possible benefit would that be for us? Hype for an unreleased product? Other companies have done this (Duke Nukem Forever, Starcraft: Ghost, the Phantom Game Console) and the community looks at those products as a joke now. Ghost was even cancelled.

I'm sorry you're having difficulty finding them, I really am, but dude, they're selling. They're online at other retailers occasionally, I know this for a fact. Newegg has sold them. ZipZoomFly has sold them. A bunch of European stores have sold them.

If you have problems with an online store not stocking them frequently enough, email the store and tell them to order more of them. I'm on your side

You want to buy our product, I want to sell you our product. We don't sell direct, so the only way to do it is through retailers. If Newegg orders 10 kits and sells out in 5 minutes, how is that our fault?

I'm not saying that's what's happening, but you have to give us the benefit of the doubt here. We're making them, selling them, and shipping them. They're in the hands of people on this very forum that bought them from those sites.

Are they hard to find? Yes. But so are PS3s and Wiis.

At least people who wanted a PS3 or a Wii could go to the store and camp out for it because they knew when it would be there. People who want good memory have to hunt for weeks and end up just buying something else out of frustration.

It's not just Corsair...it's Mushkin (although it's showing up again), G.SKill HZ too. Most of all the really high performance stuff is impossible to get hold of in the US. Like I said, I don't care one way or the other, but it is hard to find.

Don't give me crap about lookign somewhere else. Shipping is a big deal to me and I can't see paying $10+ to ship a package that is only about 2lb. Just that most of the larger e-tailers aren't showing stock on many products. Not just Corsair like I said though.

[Yellowbeard]

At least people who wanted a PS3 or a Wii could go to the store and camp out for it because they knew when it would be there. People who want good memory have to hunt for weeks and end up just buying something else out of frustration.
.

It's worth pointing out that there is NOTHING inside a PS3 or a Wii that is as rare as the Micron ICs we need to GUARANTEE 6400C3 or 8888C4. We don't build them and hope they will run at rated speeds, we guarantee they will run at rated speeds which is a major part of the reason we are able to make only a small # of them. Let's all get picnic lunches and signs and go protest at the Micron fabs.

[fhpchris]

It does when everywhere we shipped them to has sold out of them almost instantly.

I'm not trying to be smarmy, but what do you think is happening?

Do you think we released a part where we can only make like 5 of them? What possible benefit would that be for us? Hype for an unreleased product? Other companies have done this (Duke Nukem Forever, Starcraft: Ghost, the Phantom Game Console) and the community looks at those products as a joke now. Ghost was even cancelled.

I'm sorry you're having difficulty finding them, I really am, but dude, they're selling. They're online at other retailers occasionally, I know this for a fact. Newegg has sold them. ZipZoomFly has sold them. A bunch of European stores have sold them.

If you have problems with an online store not stocking them frequently enough, email the store and tell them to order more of them. I'm on your side

You want to buy our product, I want to sell you our product. We don't sell direct, so the only way to do it is through retailers. If Newegg orders 10 kits and sells out in 5 minutes, how is that our fault?

I'm not saying that's what's happening, but you have to give us the benefit of the doubt here. We're making them, selling them, and shipping them. They're in the hands of people on this very forum that bought them from those sites.

Are they hard to find? Yes. But so are PS3s and Wiis.

Why do the reps have to always come here to push product? I am all for people showing results, but I want to see results of honest-to-god purchased kits, and not every time the engineers at corsair get bored and make another revision or put different spreaders on ram.

we guarantee they will run at rated speeds which is a major part of the reason we are able to make only a small # of them.

I was very disappointed in the fact that my new Corsair stuff was DOA (funny how my 7gb+ of teamgroup ram was all perfect! Thanks Kevin!), and I get sick and tired of Corsair making the mid-grade stuff confusing for the average buyer (making Promos or Elipidia crap look like micron). I do not see Teamgroup or Mushkin making the product lines confusing so they can sell ~250$+ elipidia crap to the average joe, and I do not see reps of those companies here giving out free ram to people to push here either.

I am very excited to receive my 9136C5D tomorrow just in time for the holidays and can't wait to see how it does in my 680i, if indeed it was designed around that chipset I'll be very happy to see DDR2-1200+

Go for it, the problem with DDR 1200+ on 680I is that it is even slower than DDR798 cas 3-3-3-4 on 975 at 8m.

[Redbeard]

We realize the varety of ICs used on various modules can be confusing, which is why we PUBLISH ON OUR FORUMS which IC we use for each revision of every set of memory we sell.

Sorry for the derail, here. If anybody has any questions or comments for me, feel free to send me a PM and I'll answer it there. I'm not trying to start an "Us vs. Them" flamewar, I'm just trying to represent the company I work for in the most honest way I can, by telling the truth. I'm sorry if it came across as "pushing the product", but then again, I honestly believe we have the best, fastest memory in the world right now. So I guess I would feel pretty retarded not recommending it to people.

Let's get back to waiting for some benchmarks.

[fhpchris]

We realize the varety of ICs used on various modules can be confusing, which is why we PUBLISH ON OUR FORUMS which IC we use for each revision of every set of memory we sell.

Sorry for the derail, here. If anybody has any questions or comments for me, feel free to send me a PM and I'll answer it there. I'm not trying to start an "Us vs. Them" flamewar, I'm just trying to represent the company I work for in the most honest way I can, by telling the truth. I'm sorry if it came across as "pushing the product", but then again, I honestly believe we have the best, fastest memory in the world right now. So I guess I would feel pretty retarded not recommending it to people.

Let's get back to waiting for some benchmarks.

Yes, I know you publish that information on your forums, as well as half the people here do. The problem with people that buy middle of the road products (I know you visit OCF and [H]) is that they do not know that the 6400 version blah blah they are getting is good micron, or inferion or whatever! In some cases with online stores (like this 5400UL issue on newegg) the store depicts a good version (like 5400UL v1.2) and then when you actually get your order, find out you have 1.5.

You can see that if the most current revision of, say 6400c4, uses micron ram, and I go to newegg today and order some, what happens when I get some new revision that has Promos/Elipidia/Inferion?

That is the issue I am talking about here, and it effects innept users alot more than the ones here, as they go looking here and find that someone has some 6400c4 stuff that is micron, and they might not understand that they are getting a module that is 100% different. Changing the IC that a certain part number uses only leads to confusion, and Corsair seems to do it alot only to save money (never lowering the price when going from micron to Elipidia/Inferion/Promos either....)

If you really cared about what the users wanted, why do you not print what IC was used in the module right on the heatspreaders, or even include it in the part number to 100% avoid these problems.

[eva2000]

ftpchris, for corsair modules easiest way is to ask retailer before you buy what rev # is on corsair modules, that's what i did when i got my PC5400ULv1.2s

Guess real problem comes if more than one IC type is used on same rev # if that happens ???

[nugzo]

2 people at the egg that ordered the dominator 6400 series report that cpu-z shows this new 9136 stuff. Is it that maybe this stuff didn't bin as high so it was relabeled as 6400? or was it actually mislabeled by mistake?
http://www.newegg.com/Product/Custra...82E16820145168

I orderd a set overnight. will be here thursday. hopefully i will get the mislabeled stuff

[pentium777]

Go for it, the problem with DDR 1200+ on 680I is that it is even slower than DDR798 cas 3-3-3-4 on 975 at 8m.

Ok well since I certainly don't spend several thousands of dollars on a 30" LCD for 2560x1600 and 2x8800 GTX for running 8m PI scores I know for a that in 3d gaming synthetic and real life that at least 2 sites have proven beyond a doubt that increasing the memory speed increased FPS by 10-20 FPS going from DDR2-800 to DDR2-1200 on Conroe systems and since I need fluent gaming for 2560x1600 and eye candy turned up all the way I will enjoy that bandwidth.

[nr2134]

Ok well since I certainly don't spend several thousands of dollars on a 30" LCD for 2560x1600 and 2x8800 GTX for running 8m PI scores I know for a that in 3d gaming synthetic and real life that at least 2 sites have proven beyond a doubt that increasing the memory speed increased FPS by 10-20 FPS going from DDR2-800 to DDR2-1200 on Conroe systems and since I need fluent gaming for 2560x1600 and eye candy turned up all the way I will enjoy that bandwidth.

Links please. I went from a Crucial Ballistix 5300C4 to Corsair 6400C3 and it didnt do crap for my FPS @ higher res. The only real difference was that I could push my X6800 harder.

[Redbeard]

ftpchris, for corsair modules easiest way is to ask retailer before you buy what rev # is on corsair modules, that's what i did when i got my PC5400ULv1.2s

Guess real problem comes if more than one IC type is used on same rev # if that happens ???

The only reason for the rev #s is when components change.

So when we go from rev 1.2 to rev 1.3, something changed, most often it's an IC, but sometimes it's another component. For example, on the Pro modules we've occasionally switched brands of LEDs or something and a new rev # is created.

So if you buy 6400C4 rev1.1 and somebody else buys 6400C4 rev1.1, the parts are identical as far as ICs and components are concerned. We do not change revs for different lots of ICs though, so this month's D9GMH and next month's D9GMH will have the same lot # for the identical part, provided no other parts have changed.

[palese]

reason we are able to make only a small # of them. Let's all get picnic lunches and signs and go protest at the Micron fabs.

Yes but the problem being you send them all to England and Australia who never seem to run short

[Kunaak]

currently testing 1320 mhz, cas 5-4-4-8-TRC 12, 2T at 2.65 volts.

1300 mhz is stable easy, but theres a sharp wall at 1300 mhz, and I have 2 sets of ram that have this exact same wall... I think I am just at the wall of what this board will do.

either way, 1300 mhz isnt too hard with these boards.

http://valid.x86-secret.com/show_oc?id=140295

this is not meant to be implied that this is stable, just showing what I am playing with at the moment, and what I am currently trying to get stable.
at the moment, this appears to be about 50% stable at best, so maybe some tweaking I can get better....

[palese]

Well corsair reps you out there?? I just received my new 6400C3D and it will only boot with one stick So much for testing before they ship Yes I set the volts with the one stick and all I get is a C1 error that I have no clue to what it means.

Can some one help me understand this, I replaced my Mushkin XP2 6400 4.4.3.10 with this as I favor lower lantency, well I just did Everest benchmark and it was the worst results I have ever had, read, write, copy, latency was all below fifty %, my mushkin always at the top. What am I doing wrong?

[Omastar]

I'm not really even sure why Corsair made a PC2-9136 part. Okay, wow, it's DDR2-1142 as opposed to the DDR2-1111 part, which are both obviously built around the same binned ICs. Anyone with half a brain could overclock DDR2-1111 to DDR2-1142, unless Corsair is just eking out every last MHz they can buy doing more selective binning, and jacking up the prices disproportionately. Personally, I think the wasted hours of testing necessary to find ICs that reliably run at DDR2-1142 in addition to the DDR2-1111 part is just stupid, and could cause more hemorrhaging than profits.

But I'm not just blaming Corsair. It's a symptom of the DDR2 market right now; look at OCZ with their new hybrid-watercooled PC2-9200 part. It's just a pissing contest to see which company can wring the last few megahertz of performance out of the DDR2 standard. And it's not even impressive, since all highly binned D9GMH hits at least DDR2-1000 anyway. And those 5-5-5-15 timings aren't exactly revolutionary, either. It's just a bandwidth race now, not a timings race, because Corsair and OCZ both know finding D9 that runs DDR2-1142 or higher at 4-4-4 is going to be massively time consuming and expensive, so I guess the MHz tweaking is their last resort.

Let's bring out DDR3 already.

[pentium777]

Palese sorry to hear about the bad ram ;-( please let us know how Corsair does RMA, on high end stuff they should certainly cross-ship asap. With holidays though it sucks to have to wait.

Fortunately I have better news, received my 9136C5D, it isn't what I had completely hoped for but is what I was expecting to have for an overclock.

Memtested ram OK for a couple hours before overclocking. Was able to boot around 1300-1333 wasn't stable. Tried 1264 5-5-5-15 2.4v got system hang during test 8 (15 minute wall time)

Backed down to 1211 and was ok with 2.5v but 2.4v would error in memtest.

Running 1558FSB I switched to Linked mode and it set the ram at DDR2-1167 changed timings to 4-4-4-12 2T and left advanced to auto for all. 2.4v and 2.45v both produce lots of errors. However switching to 2.5v produces 0 errors so far after over an hour of wall time (3 passes)

So I am fairly content with 1167 @ 4-4-4-12 seeing as how I would have bought the 8888C4 @ $610 if it had been in stock, I guess I saved myself $150

BTW the ram sticks are not even warm, I do have 8x120mm fans 110cfm each and it is 74'F ambient room temp so that probably helps.

Can anyone comment if they would do 1167 @ 4-4-4-12 vs 1200 @ 5-5-5-15? I'm guessing losing the little bit of bandwidth but tighter timings is better?

[palese]

Palese sorry to hear about the bad ram ;-( please let us know how Corsair does RMA, on high end stuff they should certainly cross-ship asap. With holidays though it sucks to have to wait.

Fortunately I have better news, received my 9136C5D, it isn't what I had completely hoped for but is what I was expecting to have for an overclock.

Memtested ram OK for a couple hours before overclocking. Was able to boot around 1300-1333 wasn't stable. Tried 1264 5-5-5-15 2.4v got system hang during test 8 (15 minute wall time)

Backed down to 1211 and was ok with 2.5v but 2.4v would error in memtest.

Running 1558FSB I switched to Linked mode and it set the ram at DDR2-1167 changed timings to 4-4-4-12 2T and left advanced to auto for all. 2.4v and 2.45v both produce lots of errors. However switching to 2.5v produces 0 errors so far after over an hour of wall time (3 passes)

So I am fairly content with 1167 @ 4-4-4-12 seeing as how I would have bought the 8888C4 @ $610 if it had been in stock, I guess I saved myself $150

BTW the ram sticks are not even warm, I do have 8x120mm fans 110cfm each and it is 74'F ambient room temp so that probably helps.

Can anyone comment if they would do 1167 @ 4-4-4-12 vs 1200 @ 5-5-5-15? I'm guessing losing the little bit of bandwidth but tighter timings is better?

Glad it worked out for you, boy what are the odds of a bad c3d stick I should play the lotto

[Redbeard]

Well corsair reps you out there?? I just received my new 6400C3D and it will only boot with one stick So much for testing before they ship Yes I set the volts with the one stick and all I get is a C1 error that I have no clue to what it means.

Sorry for that, but I can assure you each stick IS tested. That doesn't mean that nothing bad happens on its way out to packaging or to you, however. C1 is usually a memory or memory size error. Definitely sounds like a bad module.

We're going to be out of the office for the US Thanksgiving Holiday for the next few days, but we'll be back next Monday, and we'll get you a replacement for those as soon as possible.

If you want to talk to me personally, PM me and I'll tell you how to set up an RMA so you can get that stuff replaced as fast as possible.

I'm not really even sure why Corsair made a PC2-9136 part. Okay, wow, it's DDR2-1142 as opposed to the DDR2-1111 part, which are both obviously built around the same binned ICs. Anyone with half a brain could overclock DDR2-1111 to DDR2-1142, unless Corsair is just eking out every last MHz they can buy doing more selective binning, and jacking up the prices disproportionately. Personally, I think the wasted hours of testing necessary to find ICs that reliably run at DDR2-1142 in addition to the DDR2-1111 part is just stupid, and could cause more hemorrhaging than profits.

The 9136 part is a C5 part, and an upscale version of the 8500C5.

The fastest part we sell is still the 8888C4DF, which overclocks faster than 9136 with the latency at CAS5 and a board that can do faster, but of course, it costs more, as well.

[zakelwe]

How far does this do 3-3-3 ?

Rather than all these high values I'd love to be able to buy some 3-3-3 RAM which will do say 450 ..or is that still a pipedream currently ? I'm not talking about the odd stick here, I mean every stick does that.

Regards

Andy

[Kunaak]

How far does this do 3-3-3 ?

Rather than all these high values I'd love to be able to buy some 3-3-3 RAM which will do say 450 ..or is that still a pipedream currently ? I'm not talking about the odd stick here, I mean every stick does that.

Regards

Andy

havent tested 333 yet.

but what you ask... that is a pipedream.

some will do cas 333 at high mhz, some wont.
you cant expect them all to, otherwise they would just start selling cas 333 900 mhz ram. luck of the draw with that kinda thing, just like with videocards, CPUs and anything else you wanna run out of spec.

[nugzo]

Redbeard.... Any comment on my previous post about the mislabeled corsair coming from newegg? http://www.xtremesystems.org/forums/...1&postcount=37
Thanks for being active and offering help here at XS.. That is great customer service in my eyes.
I almost forgot..... HAPPY THANKSGIVING XS!!