Is 1:1 ratio REALLY THAT...

Reasons why a 1:1 ratio is preferable:
-Slight improvement in performance due to the matched speeds.
-'Slow' RAM speeds which will allow higher FSB speeds.
-Better stability since the chipset strap will be clocking FSB and RAM equally.

After you've reached your desired speed you can clock your RAM at whatever speed you want, but its a lot easier and more manageable to run 1:1 while overclocking.
 
No. As for my overclocking, I have the 840 3.2Ghz 800FSB and DDR2533. The highest 1:1 ratio overclocking is at 450. That's 900FSB for cpu and 450Mhz for the memory. Now for my maximum overclocking I have achieve 4Ghz with the memory at 400Mhz. Both overclocks are stable but it would be better if I have the 4Ghz overclock on 1:1 ratio, too bad I don't have the DDR2800.
 
But in your specific case is because the memory couldnt keep up with the FSB, which would be the same case on the 650i chipset since FSB allows higher clocks than memory.
 
ANSWER THIS please!!

if ram is at 898MHz > divide 2, 449 and fsb is oc to 400MHz, then the ratio is still counted as 1:1 right? cuz 449 will round down.
 
It really depends on the quality of the Ram and other factors such as motherboard capabilities and cooling/airflow
See bellow for bandwidth comparison

1:1 @ DDR2 800


4:5 @ DDR 1000


with lesser quality ram modules 1:1 would be my vote
 
thx rwaritsdario, can u also help me on this,

just correct me.. please forgive me if i am wrong.

sample 1,
to determine the final ddr2 speed, of a E6600 oc to 400 fsb with a ratio of 4:5, take 400/4*5*2 = ddr2 6400 will run at 1000MHz.. if it's a CL4 ddr2 it might up to CL5 as 1000MHz is 200MHz above the stock 800MHz..

sample 2,
to determine the final ddr2 speed, of a E6600 oc to 333 fsb with a ratio of 1:1, take 333/1*1*2 = ddr2 6400 will run at 667Hz.. if it's a CL4 ddr2 it might down to CL3..

i tried to remember "exact divisions"
ok i know something when wrong up there. but this is the concept i got after i did some google search.
 
The looser timings allow higher speeds, and tight timings wont go very high unless you really go at it with the voltage. But this by no means has to follow a general table, every module is different and theres about 20Mhz difference within the same ICs, so making a table would be impossible.
 
For the current market the calculation is too optimistic, it is true to for Micron D9GMH/KX @2.2v but the market is so plagued with Hynix and Elpida chips among others that you cant expect that out of every 800CL4 module.
 
erm.. how do i go abt finding out wat ics are used?

i am intending to buy a corsair xms2 6400 2gb kit.. do they use micron? i heard the 1st batch was.. how do i check for the later ones?
 
RamList

I was going to recommend the buffalo but seeing it might have elpida! Looks like I need to pop the heatspreader on my stick see whats under!