Has anyone tested these yet?
http://www.crucial.com/store/partspe...KIT12864AA1067
Printable View
Has anyone tested these yet?
http://www.crucial.com/store/partspe...KIT12864AA1067
kingston and some1 else also have kits of this out
are you seriously interssted in cas7 ddr2 lol:rofl::rofl:
Interesting.
I wish there was something like this in 2 GB form though...
BTW, do you know what IC the 2x2 Ballistix kits use?
The kit: http://www.google.com/products?q=CT2...earch+Products
Single 1GB DIMM:
http://www.google.com/products?q=CT1...oducts&show=dd
http://cgi.ebay.com/Crucial-1GB-DDR2...QQcmdZViewItem (20 pieces available)
2GB DIMMs and its respective kit are also available ;)
Thread cleaned.
Don't make me step in a second time.
Just ordered the 2x1GB kit of them from Crucial. They'll be here tomorrow!
It would be interesting though to find out if we can get them to somewhere around DDR1400 on CAS7 or even 6. Some companies sell (or used to sell) 1GB sticks that did 1066 @ 2.1v with GMH chips. Seeing as how GMH is actually rated much lower, these new JKH chips might be able to clock pretty high.
That's what I'm hoping for. I just can't see Micron ditching the D9G/MH/KX for nothing. I was skimming through the data sheets last night and they seem to be more energy efficient than the Rev D chips. The specs are similar to GKX, except faster (skew settings ect.) Temp range is from 0-85c which is the same as the rev D series. Rev D series had a max voltage of 3.6V, not sure what the max voltage of the new chips are though.
I certanly hope so, may be another reason not to ungrade to ddr3 yet.Quote:
these new JKH chips might be able to clock pretty high.
i talked to some micron engineers in a conf call recently and i got quite the opposite impression ;)
but saying some GMH chips oced great so theese new chips will oc a lot as well is complete bs guys :lol:
thats like saying c2d cpus can reach 100% ocs, so all intel cpus can probably reach a 100% oc :D
yup
i got a nice collection of d9gmh saya and they not created equal lol
i got 4 sticks of ballistix "d9gmh" that are better than my d9gkx they doin 1200+ with 2.03v lol 5 months n they rockin
n my mushkin xp2 pc2-8500 never lasted more than 2 months per set lost 3 last year lol n the new sets gave me errors day 1
my kingston value d9ghm still goin strong 18 months:)
i got no faith in the binners i beleive in luck
The new chips are single sided, so essentially they can double the module capacity without making significant changes.
Everyone was mad when Micron ditched the fat bodies for the Rev D chips. It took a little while for people to realize that the Rev D chips (D9GMH/KX) were better than the fat bodies. Once that was discovered everyone was happy. ;)
The ram should be here within the hour. I'll keep you all updated. Hopefully my board detects them properly, as they do use a new SPD revision (1.3 I believe.)
ow come on, make that mailman move faster :D I'm dying here :p
NEEDS RESULTS ASAP :p:
pretty please?
Having some issues with them. The SPD settings are a little screwy because my bios hasn't had the SPD updated. My board also sucks and won't clock ram past 1168mhz no matter what. I was able to get my old Ballistix up to 1168mhz cas 4 and cas 5 on this board. These new sticks are brick walling at 1168mhz also. They don't like tight timings (as I expected, considering they're binned for cas 7.) They to be very similar to DDR3.
1168mhz 6-6-6-18 (stock voltage, 1.87V.)
1168mhz 6-5-5-15 (2.0V, haven't tested lower voltage though.)
I need to test these on a better board that can clock ram real well to get more accurate results. I suspect that these will do well over 1400mhz cas 6 and or 7 with minimal voltage with the right motherboard.
sounds interesting to me :P. I gotta look around for a kit of those, more interested in the 2gb sticks..
someone should test them on a DFI board :)
1168 @ 6-6-6 isnt bad at all, but it isnt great either. It would be nice to see these sticks on a good P35 board with 2.2-2.4 volts to see what they can really do.
Initial results are good, to be honest better than I had expected.
1168 at 2.0V isn't bad at all, from their characteristics they seem to be very similar to DDR3 indeed.
Have you tested CL5?
Well, now it's quite tempting, considering a 2GB kit isn't that expensive... But I must not buy new RAM... :cool:
Please don't post very good results or I will have to spend some money again.... :p:
the price doesnt matcht he current offerings of Ballistix 1066...
not sure why you would pay for CAS7 1066 when you can buy CAS 5 for $30 cheaper!
ALTHOUGH the 1.8v is quite nice.
someone take their D9 GMH or GKX and run it 1066 CAS7 (7-7-7-15) and see if they will run 1.8v
dont worry about overclocking right now guys just do a comparison of the 1066 CL7 to 1066 CL5 in terms of Everst RAM Speed.
see how much your performance goes down the toilet.....
although being where we are,, overclocking is a big thing.
i am rather shocked to see micron using non micron chips to be honest.
i see someone said they WERE using some other stuff and i have a set of the orange Ballistix DDR2800 and i can guarentee they are not D9 or are a really funky d9. just from their overclocking and the fact they hate anything over 2.1v
its like for using chevy parts in their cars...
4x2gb sticks would be nice but us useless as its been provne time and time again under vista64 and xp64 for anyone not running some sort of a massive server.
plus you would be hard pressed to find a system that can handle 4x2gb sticks.
whether they mfg of the mobo says it can handle 8gb of ram or not is bogus, the systems just cant handle it, be it power or northbridge or whatever...
admittedly poeple here have proven it can be done but it's not an easy task for most boards.
plus the lack of 64bit programs and STILL the lack of proper driver support or equal support as the 32bit drivers have
but anyways i am blabbering i want to see some everest benches to show bandwidth compared to 1066 ram @ CAS5
Lestat, you must not have read this thread through, and regarding your ballistix...you are way out of the loop.
These are using 'new' 78nm Micron D9 chips that are double density and in general tend not to clock as well as the older D9's that we are all familiar with. They also tend not to like tight timings.
Good luck getting 1066 Cas 7 to work on an intel board, and good luck getting the new Ballistix to actually run 1066!
I can't even get these sticks to run cas 7, not sure if it's just this board or all/most intel boards in general.
D9GMH can't run cas 6 or cas 7, it only supports cas 3, 4 and 5.
Cas 5 runs to around 1040mhz on these, but that's expected. Cas 5 on these sticks is like cas 3 on cas 5 binned IC's. It also could be an issue with the subtimings, as I have very limited manual subtimings on this board, and the spd revision the sticks use is newer than the spd my board supports.
This difference between 6-5-5-18 and 5-5-5-18 is pretty negligible. I'll get some screenies in a few of some benches.
I'm stunned, no difference between 6-5-5-18 and 5-5-5-18!
*note if you're wondering about why the SPD says PC-8800 555mhz, it's because of the difference in SPD revisions. The SPD byte value on SPD 13 of '18' (1.8ns) is 533mhz, on the board SPD rev 12, the value '18' is 1.8ns (555mhz.) There is no 533mhz spd value on spd rev 12 (or 1.2.)
http://img42.imagevenue.com/loc792/t..._122_792lo.JPGhttp://img180.imagevenue.com/loc85/t...h_122_85lo.JPGhttp://img225.imagevenue.com/loc139/..._122_139lo.JPG
i have been out of the loop man, i been real busy at work and really haven't had time or energy to do a lot of research.
so if my comments seem uneducated (about my own hardware) then you are correct.
http://www.crucial.com/store/partspe...KIT25664AA1067
2x2gb 1066 CL7 1.8v
WAYYYYYYY too expensive.....
The ram in my sig, as well as another set of 512mb Ballistix are as good as dead. They need atleast 2.3V to boot and they won't run faster than 667mhz 5-5-5-15. It takes about an hour of fiddling just to get them to post.
They were great before though, hit 1000mhz cas 3 with the set in my sig, 2.81V.
Theres a market for pretty much everything ;). Especially, if your motherboard cant run 4 sticks properly, guess 2gb modules are the only way to get 4 gb :p And there should be quite a difference between 2 and 4gigs on 64bit no? Powerchips are all nice and cool, but if these offer similar performance, why not :).
(other than the price atm :X)
Hmm is it just me or the screenies are identical? Can it be you screenshot the 5-5-5-18 2 times? or is the proxy at work being weird oO.
Anyway, is the difference in bandwidth between 5-5-5-18 and 6-5-5-18 supposed to be big? Cant try out right now.. Was under the impression PL is way more of a factor.. Is a PL of 6 physically possible at 6-x-x-x :x. Never had the time to tweak my mobo much :(
PS sorry if I make some weird conclusion, am pretty new to the developments over the last few years :D
Guys...are you talking about DDR2 here? Because I had the feeling that this thread is about DDR3......:)
Working 4x2Gb has absolutely no point, so that would not be the reason for which they produced these new chips. So...I must say....Unless they run >1500 CAS 7....they are pretty much useless.....
yeah, definately... for some batches gmh and gkx are almost identical when you oc them, for some batches theres a big difference and all chips are good... some batches oc better with cas4 than with cas5, lol...
the last batches all clocked very well with cas5 iirc
the older batches oced better with cas4
at least thats the experience i made...
well, you wont believe it, but micron isnt as aware of how well their chips clock as we are... :D their priority is after all the OEM business which means they need chips that have good yields at the jedec timings and speeds, with as low voltage and power consumption as possible...
what makes micron a good ocing chips mfg is that they always aim to reach the highest possible jedec spec or surpass it though, while most other chips makers usually only want to make chips that barely run the volume jedec specs like ddr2 667 555 for as cheap as possible.
hutch, oh come on!
you bought high speed mem and try to push it on a freq limited board? :P
christ, all this waiting for your results and then this, meh!
i want my money back! :lol:
hope you can find a decent board asap :D
In the first screenshot I didn't include the benchmark for 6-5-5-18. The 2nd one included the benchmark and the 3rd one included the benchmark with timings set at 5-5-5-18.
Going from PL 8 to PL 7 yielded a boost of 250mb/s on the read score and -3ns on the latency. I hadn't tweaked the timings at all there. The PL can probably be brought down to 5 or 6 at this FSB.
1111mhz 6-5-5-18 1.87V 463mhz FSB 5:6 divider on E6400 (333mhz bsel mod.)
http://img139.imagevenue.com/loc1112...122_1112lo.JPG
nice, i'm waiting for your results
really excited to see these new chips, hope we can get some 1400+ clocks =D
good to hear alex!
try not to blow up your board until the weekend then :D
no benching for you all week! :nono: :D
We both got too tied up to do anything with them today, but I guess we'll have to try for another day since you guys are clammering for some more numbers.
hi
ive got a kit of these crucials too. made some quick test on a gigabyte p35 ds3 and the results are weird. no matter whats the voltage from 1.72 to 2.25 i always get stuck at (2:3 ratio) 1142 cl7. cl6 is slightly weaker, but increased voltage wont help either.:shrug:
any suggestions? i was expecting these modules to go crazy even on cl5, but it has been quite a disappointment so far. will try different ratios, i hope that will help. d9gxx chips liked 2:3, but these are not only a bit different from those... maybe they like something else.
the spd is useless for this ram :) since its rated 1110 cl5 1.8v, of course it wont even boot up at these settings.
so whats happening with these chips :)
I'll be testing them on the new Biostar P45 board once it gets here. Ordered it yesterday from newegg.