What memory
- Thread starter Modcon
- Start date
And where did you pull that answer from ??? did you throw a few answers into a hat and pick one ?
and about the OP topic, go for pc6400 DDR2 its not the name of the company that matters when your buying ram, its whats under neath those heatspreaders what matters most.
you want to be looking for *MICRON D9 ICs*
here is a tip, this is the stuff i just bought though i bought directly from Crucial becuse oCUK dont seem to have it in stock at the moment.
Cost me £120 for 4Gig of it.
Crucial Balistix Tracers PC6400 DDR2 4x1
When you install that stuff it has a stock speed of 4.4.4.12 T2 @400mhz (800) but if you open CPUZ you will notice it has sli ready profiles EPP.
the stuff can do 5.5.5.15 @500mhz (1000)
but thats just what the stuff can do when you stick to the stock profiles
------------------------------------
with all sticks in my 680I its just the same board as a 780I the only diffrence is PCIE 2.0 and 3way sli...
4.4.4.12 T1 @450mhz (950) with my quad at 8x1800FSB 2.2vdimm 3.6ghz
4.3.3.9 T2 @450mhz (950) with my quad at 8x1800FSB 2.2vdimm 3.6ghz
now i would hate to think of the speed these could hit when i change the timeings to 5.5.5.15 T2 then up the Mhz
see i just got this stuff yesterday seems like great stuff and the flashing lights on them just add to the shear pleasure of useing them and did i forget to say they use MICRON D9 Ics...
anyway i will leave it up to you to decide...
Don't bother with flashy high speed memory or even bother with memory overclocking (unless you have a quad core at 3.6+ gig) there are no appreciable real world gains... get some PC2-6400 either OCuk value or OCZ platinum rev 2 imo - unfortunatly there isn't ant crucial 6400 value RAM on OCuk, if you really must overclock then get some D9 based RAM like the g.skills 6400HZ or Crucial Ballistix...
Thats some fantastic results with 1T ViruS2k... tbh a bit to good to believe they are 100% stable... with the Asus Striker Extreme (horrible, horrible board) I could not get higher than 850MHz 4-4-4-12 1T max with any D9GMH based chip regardless of how much voltage I gave it... it would boot upto 960MHz but it wouldn't get through 3 passes of memtest without error.
Heres an idea of what you can get with 5-5-5- timings on 2T
http://aten-hosted.com/images/spi3.jpg - could do 5-5-5-15 with more voltage but I found I could drop the voltage quite a bit with 5-5-5-18 with less than 0.5% performance drop. Even so they died after 18 months at those kinda speeds/voltage.
Heres an idea of what you can get with 5-5-5- timings on 2T
http://aten-hosted.com/images/spi3.jpg - could do 5-5-5-15 with more voltage but I found I could drop the voltage quite a bit with 5-5-5-18 with less than 0.5% performance drop. Even so they died after 18 months at those kinda speeds/voltage.
You need proof ???
im currently still tweaking and stuff but all this you see on this image is true.
im working with the guys over at EVGA forums on tweaking them to there max.... here is a image i took tonight, and before i took that image i primed my system 8+ hours stable with this.
and memtested no errors.
see for yourself.
btw check out my CPU temps after doing the vdoop mod on my motherboard its made a world of diffrence to temps aswell as voltage !!
all on air.
im currently still tweaking and stuff but all this you see on this image is true.
im working with the guys over at EVGA forums on tweaking them to there max.... here is a image i took tonight, and before i took that image i primed my system 8+ hours stable with this.
and memtested no errors.
see for yourself.
btw check out my CPU temps
after doing the vdoop mod on my motherboard its made a world of diffrence to temps aswell as voltage !!all on air.
As you can see in the picture there is something i just havent figured out, why im getting low bandwith efficiencey....
for some reason im unsure at the moment what the cause is of that as yet lol
with that sorted i should be hitting about 5% more speeds in memory.
for some reason im unsure at the moment what the cause is of that as yet lol
with that sorted i should be hitting about 5% more speeds in memory.
You'll always get that low bandwidth efficency error - iirc it compares the theoretical maximum bus bandwidth with your benchmark figures...
I wasn't looking for proof - I'm not sure what the EVGA board is or isn't capable of - its just a lot higher than the average with a 680i board and those kinda wildcard results are always gonna raise eyebrows...
I wasn't looking for proof - I'm not sure what the EVGA board is or isn't capable of - its just a lot higher than the average with a 680i board and those kinda wildcard results are always gonna raise eyebrows...