my Q9300 is with DDR2, DRAM : FSB = 1.60 : 1 (or Memory Multiplier 3.20x CPU FSB 333), the DDR2 memory timing was no tweak.
the QX6700 is also using a P35 board, but it's DDR3, DRAM : FSB = 2.00 : 1 (or Memory Multiplier 4.00x CPU FSB 266).
If i run the Q9300 with DDR3 and use the same 2.00 : 1 (or Memory Multiplier 4.00x CPU FSB 333),
then the memory would run at DDR3-1333, and i should get a better results (all around).
However, these benchmarks are synthetic, ie. the results show only the theoretical (maximum) performance of the system. In contrast to application tests, synthetic benchmarks do not tend to reflect the "real world" performance of the computer.
the benchmarks provide quick and easy comparison between computer states, e.g. when certain parameters (CPU clock speed, memory timings, etc) change in system configuration.
__________________
Q9300 @ 500FSB
Q9300 @ 3.60GHz
4x1GB DDR2-1333 (PC2-10666)
-------------------------------------------------
Don't worry overly much about whether a specific KB article exists for your version of Windows.
Standard practice is to grab whatever's available and apply the principles.
(I grabbed from the web. Written by someone, somewhere on the web)