This memory is based on Winbond UTT -> handpicked 2-2-2 @DDR400. (BH5)Quote:
Originally Posted by gsan
B6U808 = Brainpower PCB and Double sided 16x 32MB.
Printable View
This memory is based on Winbond UTT -> handpicked 2-2-2 @DDR400. (BH5)Quote:
Originally Posted by gsan
B6U808 = Brainpower PCB and Double sided 16x 32MB.
@GSkill:
I see some of your modules are spec'd to 1T others 2T...
This cought my eye on the AMD Processor Support Forum. (see the 3rd quote in 5th post) :confused:
Quote:
Overall, rating a module as 1T is either false advertising or needs to be excused as blatant ignorance as all unbuffered modules are capable of a 1T CMD rate up to four banks per channel, beyond which chipset limitations come into play.
thats true, only instead 'chipset limitations' i'd say 'memory controller limitations'. but thats commonn sense these days.
Technically that is correct. Command rate (CMD) is memorycontroller (chipset). Thats why you can run 2T rated gskill modules at 1T if you have a descend memorycontroller.
Tested by myself.
News
All G.SKILL Extreme Series DDR1 with buidl 0508xxxxxx will have Heatspreaders.
The chips will be still the same. (TCCD507 and later)
http://xtremesystems.org/forums/atta...id=35935&stc=1
my gskill 2x512MB GH seem cant perform well on NF7-S :(
even 230 @ 3.4v cant stable.
Their are some common problems when running with NF2 S-A platform. It is not the Gskills. So will see that single channel mode will go much faster.Quote:
Originally Posted by gsan
Make sure you run 1T, or if the onchip-controller don't want run low timings @230+, try 2T commandrate.
Personally, i don't like heat spreaders. I actually liked GSkill rams bare where I can blow a 80mm fan on it and the warranty stickers are intact. Heat spreaders to me are more like a heat trap.
FC-Review at XBitlabs: http://www.xbitlabs.com/articles/mem...ay/gskill.html
I agree with that! If you want overclock you must put a 80mm+ above the memory, then the spreaders will stay cool.Quote:
Originally Posted by Spot
Well just tried the new LEs with heatspreaders and they aint all that impressive. At stock, they do their rated settings at 2.8v (275MHz 2.5-3-3-6) but gg up to 300Mhz, only could get 2.5-4-4-8 Memtest stable at 3.1v. This is comparing to my FCs (TCC5) which are now memtest stable at 2.5-3-3-6 300MHz @ 3.0v.Quote:
Originally Posted by kakaroto
Got a 60mm over the sticks and even with it on there, the sticks really arnt getting hot. I truly question the usage of the heatspreaders considering the thermal pad appears to be only 3~4mm in width and hardly covers the base, (gg with what Spot mentioned above), its like a huge heat trap more than anything. Case in point, the spreaders should be getting at least slightly warm if there was decent thermal conduction gg on.
Quite dissappointed unfortunately with this pair of LEs.
Glad you agree. Do you think GSkill will do away with the heatspreaders here on? It adds to an extra process and extra cost for GSkill and not only it doesn't help at all, it actually impedes performance.Quote:
Originally Posted by kakaroto
i have FF&GH
very good Product :banana:
LE's generally do not like that much voltage, you may get better results with 2.9 or less. Did the RAM scale with voltage at all past 2.8?
Quote:
Originally Posted by Sin22
thanks for te info.
love the Heatspreaders, they make the rams look very sexy :toast:
Not a fan of the heatspreader. I have always preferred my memory the same way I like my women ... naked and hot. ;)Quote:
Originally Posted by cupra
When is someone going to get some F1-3200USU2-2GBHS for testing? Need to know the type to Samsung IC used on these. If its UCCC, its may be a mediorce performer. I'm hoping GSkill has an ace up its sleeve with this 2x1gig set. :D
LOL! :lol2: :toast2:Quote:
Originally Posted by gundamit
UCCC is capable of very high speeds.Quote:
Originally Posted by gundamit
wow nice post
looks like some good RAM
Will really like to know how UCCCs perform. 2x1GB is the next thing.Quote:
Originally Posted by Vapor
2.8v could only get it stable at rated freqs. 275MHz 2.5-3-3-6Quote:
Originally Posted by Eldonko
At that voltage it could only get up to 280MHz 2.5-3-3-6, afterwhich voltage bumps to get it stable.
And to get it stable at 300MHz, it needed 3.1v & and loose timings of 2.5-4-4-8. Nothing less would do it unfortunately.
Burning in would help Im sure, but not that much.
The thing is, its one of two things, A) got crap o/cing sticks (seeing my luck these days with RAM I wouldnt be surprised) B) The heatspreaders are also limiting the o/cability by inducing a heat trap due to the small TIM pad and hardly any contact with the spreaders themselves. Its been my experience that removing heat spreaders (even decent ones with nearly full sized TIM pads) does tend to improve o/cing results significantly.
Does anyone have any idea what I can expect from 2x512 G.Skill FF's on a DFI nF3 Ultra-D?
Thanks!
The same as on on the nF4 series.Quote:
Originally Posted by ixtapalapaquetl
The only difference between those two boards are the Graphics Bus and the chipset. nF3 is also capable of doing 300Mhz HTT.
do you have recomended settings for F1-3200DSU2-1GBLA DDR600 running on intel p4c systems.
i had problems getting past 280 with any timings.
Quote:
Originally Posted by kakaroto