*HOT* | Corsair DOMINATOR GT | 6GB (3x2GB) | DDR3-2000 7-8-7-20 @ 1.65V | *HOT*

[jagt]

What motherboard are you using? Damnit, my RMA will be here tomorrow or thuesday. I'm not sure if I dare using them at all now! :P Looks like we have the "new" D9 GMH's...

And the RMA arrived. Looks like this kit is... sh*t compared to the other kit I had. Won't even POST at 2000MHz 7-8-7, whilst the other kit was stable at that speed at stock voltage. When it does post the system can only find 4GB, so I assume on of the sticks is just crap. Errors out in memtest at 2GHz 8-8-8 too. Testing now with a slight voltage bump

Oh yeah, and my system only posts half of the time. The other times it will be stuck in a post-loop where the only solution is to clear CMOS, even though the settings proved stable before rebooting.

[saaya]

"it has no effect whatsoever on a possible rma"

ROFL are you high or something?!?

geez, calm down dinos...
what i mean is posting something here has no effect on an rma...
even if the nickname somebody has on hir forum here is similar to his real name, and most arent, then he would still claim its a diferent person if corsair tech support asked him about this forum post...

even if i post here and say i killed my gskill memory with 2.4v and then rma it, and the tech support tells me, hey, theres a sascha aka saaya on xs who said he killed his mem with 2.4v ... ill say its not me, must be another guy... so then what? they will say i have the same first and last name and am from the same country? and? do you know how many people in germany have the same first and last name as me? i know of at least 11 cause i once looked it up when i worked for the german telecom

and even if there is only one guy with my first and last name... so what?
do you think any manufacturer can honestly afford to risk refusing a legit rma cause they THINK the guy MIGHT have killed his product by misusing it? i dont know how this is dont in australia, but in germany youll get sued if you refuse an rma and cant prove that you have a reason to refuse it...

lets say X created an rma for the mem he killed with high vdimm, then he might be careful and wont post on the forums that he killed his mem with high vdimm... but would he make an effort and post everywhere that his mem died at stock vdimm? why? to risk pssing off the manufacturer?

you seem to think all those guys who posted here only said their mem died at low vdimm cause they are affraid they wont be able to rma it, and actually they killed it with high vdimm... i really dont think so...

[jagt]

Bah, this new kits blows. It doesn't wanna run anything above the stock settings! Night/day difference between this and my last set that did this just fine:



If I understand the serial number correctly, this kit is one week newer than the previois kit I had that died (0918 vs. 0917). They act like they are different IC's. Perhaps this is of the type that overclocks like crap but lasts longer than two days? :P

[bluehaze]

It's probably different IC's now. I imagine Corsair has been getting tons of this ram back for RMA and was probably well aware of the problem long before any of us. I also notice the 1866 stuff is getting much cheaper, it's only 234.00 now on Newegg.

[jagt]

It still says Rev 2.1 on the sticker. Don't they usally change revision when they start using a new IC? I can't even pass LinX at the stock 1866MHz 7-8-7 on this kit. It's utter and complete . I'll start testing each stick individually now.

[bluehaze]

It still says Rev 2.1 on the sticker. Don't they usally change revision when they start using a new IC? I can't even pass LinX at the stock 1866MHz 7-8-7 on this kit. It's utter and complete . I'll start testing each stick individually now.

Yea, I would think so...not sure how it works. I am really hoping to get a refund for mine though. Will buy some 2000mhz later once all this gets worked out.

[jagt]

Yeah, that sounds like a good plan. I would do the same but I doubt my e-tailer will allow me a refund. You pay a premium and this is what you get.

[tekjunkie]

I got my replacement GTs and so far so good.



I really REALLY hope that this kit does not die on me. If it dies on me, O well, $500 down the drain. I am sick and tired of RMAing.

[dinos22]

i dont think that 7-7-7-x is stock

[cstkl1]

gave up a while on 2100 4.4ghz

now testing running 2k 7-7-6-20 tRFC 60 vdimm 1.6v

so far i found linx is easy
prime is hard say 8-12 hours especially after 9 hours
cause pwm frequency also comes into play.
each vtt has different optimal pwm frequency
dram speeds are depended on the dram pwm frequency also.

example 1.36 perfect at 180khz
1.4 prefers 153khz
1.44 prefers 126khz
dram 2k prefers 180khz
2050 prefers 207khz etc

[cstkl1]

i dont think that 7-7-7-x is stock

well classified
testing on one board everest result is almost = to dmm read.
eleet is way off on load
when vdroop is disabled.



test setup 950
adata xpg 2133
classified + bits full coverage + mosfet+ht3+feser 360
corsair hx1000

[eva2000]

so eleet vs dmm

vdimm 1.64v vs 1.68v real
cpu vtt 1.475v vs 1.55v real
ioh vcore 1.1v vs 1.13v real
vcore 1.45625v vs 1.47v real

???

[SAE]

Eleet likes to show more Vtt than there really is:

BIOS set -> Eleet -> DMM

1.425Vtt -> 1.52 -> 1.4000
1.62Vdimm -> 1.65V -> 1.6440




Mine are 0918 too, they don't suck and I luv and want to marry them.

[jagt]

I will be RMA'ing my kit again as they refuse to be stable even at stock settings. I have underclocked them to 1600MHz 7-7-7 now, and they won't even POST at 6-7-6! Talk about rubbish kit. Oh well, 3rd time's the charm.

[drnip]

so eleet vs dmm

vdimm 1.64v vs 1.68v real
cpu vtt 1.475v vs 1.55v real
ioh vcore 1.1v vs 1.13v real
vcore 1.45625v vs 1.47v real

???

Yeah I noticed on my Classy that the voltage tab of Eleet and BIOS is backwards or at least to me. A setting of 1.65v DIMM in BIOS shows 1.65v in the Eleet voltage tab but reports 1.71v on the monitoring tab of Eleet as well as other monitoring utilities? Same goes for other voltages. Usually in my experiences what is reported is lower than what is set. My vcore is normal though. It reports lower than what is set.

[cstkl1]

Eleet likes to show more Vtt than there really is:

BIOS set -> Eleet -> DMM

1.425Vtt -> 1.52 -> 1.4000
1.62Vdimm -> 1.65V -> 1.6440




Mine are 0918 too, they don't suck and I luv and want to marry them.

interesting
i find the problem is only on load , it goes up on my dmm and when vdroop is disable. talking about vtt and vcore.

are u using mch 1600 strap .
hmm see eva i was right. told jimmy already. that the 2050 on dfi on 205. all the bootup clock skews are wrong.
from his rtl i can already tell.

[SAE]

I am using strap 1600, yep.

My VTT and Vdimm droop under load, but it's minimal. I am using max PWM frequencies. The VTT being shown .1V higher that it really is, seems to be the standard with the Classified. It's also setting the VTT .025V lower than it's been set up in BIOS/Eleet.

Vcore droops if LLC is off (that's what I am using), but switching Vdroop off pushes Vcore under load. Makes perfect sense in my eyes, mate.

[cstkl1]

I am using strap 1600, yep.

My VTT and Vdimm droop under load, but it's minimal. I am using max PWM frequencies. The VTT being shown .1V higher that it really is, seems to be the standard with the Classified. It's also setting the VTT .025V lower than it's been set up in BIOS/Eleet.

Vcore droops if LLC is off (that's what I am using), but switching Vdroop off pushes Vcore under load. Makes perfect sense in my eyes, mate.

actually u made me realize something about the classy
ure mem skews.which affects the rtl's
ures is same as my 2k. setting on dfi..
its impossible to rtl 53 on dfi for 2050.

going back to test the dfi again.

[bluehaze]

actually u made me realize something about the classy
ure mem skews.which affects the rtl's
ures is same as my 2k. setting on dfi..
its impossible to rtl 53 on dfi for 2050.

going back to test the dfi again.

I don't understand clock skews if that is what you are talking about? But I am curious about what you say because I notice that my Classified has lesser performance than other boards at the same clock speeds and have been tweaking timings etc... to try and close that gap. Basically in 3dmark Vantage I lose about 1000 3dmarks compared to a UD5 at same speeds and no amount of tweaking has made up for it yet. Sorry if this is a noobish question but I have not yet messed with any of the skew settings, does this affect performance of the board?

[cstkl1]

I don't understand clock skews if that is what you are talking about? But I am curious about what you say because I notice that my Classified has lesser performance than other boards at the same clock speeds and have been tweaking timings etc... to try and close that gap. Basically in 3dmark Vantage I lose about 1000 3dmarks compared to a UD5 at same speeds and no amount of tweaking has made up for it yet. Sorry if this is a noobish question but I have not yet messed with any of the skew settings, does this affect performance of the board?

that could be any factor
from ure os interaction with eist
to many things
or qpi strengths

but the dram bandwidth is the about the same between dfi and evga .
mem skews that i am talking about is a internal setting in the bios
for dfi it is set via bootup clocks which are tweaked by their engineers
evga is via mch strap.

rtl 53 at 2050 c7 is really unreal and awesome dfi cannot do this atm with
their current bios afaik.

[eva2000]

that could be any factor
from ure os interaction with eist
to many things
or qpi strengths

but the dram bandwidth is the about the same between dfi and evga .
mem skews that i am talking about is a internal setting in the bios
for dfi it is set via bootup clocks which are tweaked by their engineers
evga is via mch strap.

rtl 53 at 2050 c7 is really unreal and awesome dfi cannot do this atm with
their current bios afaik.

pretty close hehe

RTL 55





RTL 56

[cstkl1]

nice eva
click my sig and u will see 100 linx 4.3ghz 1050 C7 rtl 54.

53.. is not possible on the dfi atleast for me for 1050.

and willing to bet again his rtl on 1050 = 1000mhz C7.

[SAE]

I could do RTL 53-56-58 @ anything up to 1055 7-8-7-20 1T. But I feel there's coming some kinda instability then. I am still struggling to get anything over 21x206 stable with 10x RAM 7-7-7 or 7-8-7... might be a crappy IMC then.

[bluehaze]

that could be any factor
from ure os interaction with eist
to many things
or qpi strengths

but the dram bandwidth is the about the same between dfi and evga .
mem skews that i am talking about is a internal setting in the bios
for dfi it is set via bootup clocks which are tweaked by their engineers
evga is via mch strap.

rtl 53 at 2050 c7 is really unreal and awesome dfi cannot do this atm with
their current bios afaik.

K, Will play with rtl! Haven't messed with it much mostly just standard timings and straps because I don't understand enough about the others. Trying to break 30k Vantage single gtx 295...so close right now at 29,700.

[saaya]

I don't understand clock skews if that is what you are talking about? But I am curious about what you say because I notice that my Classified has lesser performance than other boards at the same clock speeds and have been tweaking timings etc... to try and close that gap. Basically in 3dmark Vantage I lose about 1000 3dmarks compared to a UD5 at same speeds and no amount of tweaking has made up for it yet. Sorry if this is a noobish question but I have not yet messed with any of the skew settings, does this affect performance of the board?

check the BR review on xcpus.com, they compared it to the classified and saw the same weird perf drops, it only seems to affect some memory dividers... it could be that you have to manually adjust the mch straps for the diferent memory dividers... i dont know why clock for clock the classified is slower, but i wouldnt be surprised if its caused by some relaxed timings they use to get higher bclock/qpi speeds? i dont know if you can actually adjust this and get back to tight timings and good clock per clock performance for people who really dont need high bclocks...

that could be any factor
from ure os interaction with eist
to many things
or qpi strengths

but the dram bandwidth is the about the same between dfi and evga .
mem skews that i am talking about is a internal setting in the bios
for dfi it is set via bootup clocks which are tweaked by their engineers
evga is via mch strap.

rtl 53 at 2050 c7 is really unreal and awesome dfi cannot do this atm with
their current bios afaik.

yeah i thought so too, but i dont think its turbo since it affects diferent mem speeds... some speeds are fine, some are notably slow

thats odd about dfi btw, are you saying they decide when to chose what strap and dont allow users to adjust it manually in bios? thats kinda weird for dfi...

I could do RTL 53-56-58 @ anything up to 1055 7-8-7-20 1T. But I feel there's coming some kinda instability then. I am still struggling to get anything over 21x206 stable with 10x RAM 7-7-7 or 7-8-7... might be a crappy IMC then.

but with lower memory multiplier and same uncore multiplier you can get higher? try more vtt and... i dont know how they are called on the evga board, but increase the x58 voltages, theres a 1.1 and a 1.5v voltage, increase both and see if it helps... can you get higher with cas8? i was quite surprised when raja showed how much memory can hold you back bclock wise... he switched from some good samsung to some good elpida sticks and it gave him 10mhz higher bclocks or even more iirc... and he couldnt get those even with relaxed timings before...

its weird, and i dont really get why... but several people had the same experience, they couldnt get high bclocks even with high vdimm, vtt and loose timings, then switched to another set of mem and could get noticably higher bclocks then...

whats the highest bclock you can boot at SAE?
your on phase change right?