Mine are green, retail.
Printable View
i can confirm that the ocz blades has no xmp profiles.
was told by the tech in taiwan when i enquired few months ago.
not sure about the latest batches
since we are on the topic of ocz's
what chipsets are the flex 1866 Cl7?? and 1600 CL6??
also hows ure scaling like for dram mhz to vdimm to vtt..
i find all three very linear.
also any chance of doing a DFI UT vs EVGA Classified on Mem Stability Benching on the Hypers??
think a lot of ppl would be interested on this.
Mine, black, retail
interesting, no big deal really...i would have just thought that with such a high end product they would try and make sure they all got black PCB's, from strictly an aesthetic point of view. clearly some are coming with black, but others green.
very nice sticks
Hmm,really? The new Beta bios seems to solve a lot of problems and myself was able to do 200x20 with the memory 2:10 on air(TRUE black push/pull) with my W3520 and the patriot 2K cl8 kits that have elipdas on them, on it.Did you try the board with 04/28 or the 05/08 betas?
green is most likely ref design or the default pcb for that chip from brainpower or whoever makes the pcb, black is usually custom...
from all pcb colors black has the worst yields for some reason, its somehow dificult to paint pcbs black without hurting quality... dont ask me why :confused:
so green is the default usually, and everything else is custom...
i got green. this sticks are hard to oc...above 2050mhz...hmmm sub-timings are great though at default frequencies...
could very well be right...although it use to play nice, then after the Classified showed up, it seemed to get all crusty. jealousy is my guess.
now do you mean 24/7 stable hard to clock, or just single 32M hard to clock? up to and over 1000MHz for 24/7 stability like Prime Blend, HCI, etc. is all about the IMC. trust me i know, i have two CPU's with terrible IMC's. unlike George above who seems to get killer IMC's (could just be his CPU/MB/Skill combo), not all CPU's can run 8x32M at damn near 1100MHz.
personally dont think 1100mhz is hard
the prob is getting the cpu speed example 4.6ghz with the ram 2200 is hard..
will know next week.
got 3 nice kits to test it out in the classy
My first set of Blades had to be RMA'ed as it needed about 1.7Vdimm to go 1000 7-8-7-20 stable. Those were the ones with a black pcb.
I then got a new kit with green pcbs as a replacement. They were running nicely - been memtesting them at 1.636Vdimm and 1.34-1.36Vtt (DMM) yesterday successfully. Luckily I switched the rig off to go to bed. Today I wanted to stress them w/LinX and Prime x64... but I got no chance as nothing was stable anymore.
-> Finally: I found out stick #3 crapped out and now won't boot anymore. Seems that was one of the kits with flaky ICs. :(
Man, I really don't know what is up with the Elipda kits dying. I have never seen so many deaths of memory. I mean even after so much abuse from a lot of people a lot of kits do not die like this. I personally had 4 kits which had Elpidas in which 1 of the sticks was bad or went bad after some moderate use. I mean I am not talking about something like 1.9 or 2.0 vDimm, this is all with stocks volts or a bit over stock. Sorry for the rant guys, I just hate RMAing and these kits dying is driving me nuts.
Second RMA for me... nothing more than 1.66V given, testing done using Memtest86+ 2.11 and voltages measured via DMM.
^^^Had a similar problem with my blades
First two sets died. One while running my first hyperpi run at only 1600mhz 7-7-7. Second after about 3 days of use and still only had them at 1800mhz at the time. Both times the stick in the third slot failed, so I replaced the board as well and now I've been running for about a month with no issues on this kit.
I am very curious to see that they have gren pcb's, not black. All 3 of my kits were/are black.
The DFI UT board is having some major problems getting elpida memory to work with a D0 revision on sticks not named "Dominator GT." I know Corsair used a proprietary PCB for their early revisions and those seem to be the ones that are working ok now. But all the other sticks are having some problems. And elpida benchmark scores clock for clock with the D0 processors vs. C0/C1's have some strange differences.
Right now, I can run the specs shown in my sig and thats about it. For some reason if I increase vtt past 1.23 set in BIOS, it is unstable. If I set VTT to 1.31 higher it simply will not post with C1 or CF. With a previous C0 2000 7-8-7-20-1T was not a problem with 1.65vdimm and 1.41vtt.
I'd be curious to see if the green pcb version works any better. Might give OCZ a call and see if they'll swap em out for me. Luckily i live about 30 min from their USA headquarters and they do will call RMA's :)
Very nice kit :up: i just slobber:rofl:
if i need buy some good DDR3 i will realy thinking abouth this
thank you for little review
after seeing Raja post his silly 8-9-8 results with the PC3-17000 kit, i thought i would see what these sticks do at 8-9-8...not bad for 1.65v :D
http://3oh6photos.com/oc_report/ram/.../8-9-8-x-1.png
1112mhz 8-9-8-20 at 1.65v :cool:
Guess we're all so focused on 7-8-7-20 we forgot about looser timings as such :)