MSI X58 Platnium & 12gb DDR3/1600??

[ciro]

does ASUS board with that 'extra options' really help? Im feeling I should gone for the full sized ATX board and larger case instead of the mATX board in square shape, Northbridge reads 60 Celsius and cant do much as its a tiny case
anyway we are on same boat, finger crossed for any solution

[genetix]

Be nice if someone from Gskill would post the full tiimings needed for this... Got the 9-9-9-24 down, need the rest....

yeah, indeed have to agree you on there, but since things ain't so. Dump me the values you need I can see what I can muster.

Basically every single value on board can be calculated accordingly on clock speed to memory internal speed readings from actual memory profile on first 256KB of memory with right tools.

[screwtech02]

Wow, just put win7 Ultimate on my box, now it seems i can run the 4/8 DDR3/1600 without a hitch. Everest bandwidth test passes with NO issues. Who woulda thunk??

Oh, and to any Gskill techs out there, i'd still like to have the secondary timings for the DDR3/1600 if you would be so kind to post em....

[screwtech02]

Genetix, since no-one else seems to be able to help me, what do you need from me to get the correct primary/secondary timings???

[genetix]

Well, wouldn't mind an everest screenshot of chipset page of MCH, if that shows the current timings and voltage you are running controller, atm. Basically, need to know what values you can adjust on secondary/third phase delays. Perhaps I can take a look to Motherboard manual somewhere should be most accurate. However, most of all these things bases on testing (most of the restrictions what you cannot use are based on Motherboard Manufacturer restrictions not what RAM actually could do), but I've tested RAM for past (hell lost count on years), but still I can only point you the ones which might have ideal improvement on speed & flexibility(if there is any in the end, lol).

Also we need to understand that speed of the memory and how it actually works, if we do some tweaking are completely in syncronous state. Example: DDR2 at 1066-1333Mhz will be slower than DDR2 at 1:1 900Mhz in every single case except when we lower it from 900Mhz with FSB. On stock FSB like 333Mhz, however, this status changes and higher ratios are more than the CPU->memory throughput.

So, as you can now clearly see this has totally 2 sides in the matter all depending how you want your QPI/FSB to run and how you want the memory to run.

Edit #01:

Also to explain the idea on general timings(I don't even try to offend your intelligence) to the secondary delays. everything else except tRD affects whole controller (Read and Write and Read/write) speed of the memory. To get REAL speed you need less degraded tRD to generate ideal 'full controller speed' -> read ONLY speed improvement to make read speed to be closer/better sync with how RAM functions while on use. This is where the hardest part comes in as different Ratio of memory has an different +Xt to tRD. on 1:1 max its +3T on 1:1 at strapped below max is +1-2T on any ratio with more than 1 will be +1T(this is still reducable to pull-down, but it won't be as fast even as 1:1 with +3T as it doesn't take advantage of Processor L2/L3 caches then).



Truly sorry for a novel there. Sometimes it just takes a lot to get the idea behind something explained.