Request opinions: Effect on Yorkfield OC of having 4-8GB RAM
I'm planning a high-end WinXP rig in Mar-Apr, aimed at a mix of video and audio encoding, and gaming. It will have either the QX9650 or the Q9550, and at least 4GB of DDR2. The RAM will probably be 2x2GB sticks so I can later go to 8GB if I ever move to Vista64 (for big video files). I also plan on a high-end Nvidia graphics card.
I have to consider the ~$500 difference between the QX9650 and the Q9550. If the upcoming C1 stepping QX9650 is likely to OC to 400MHz more than the initial-release Q9550, I could justify the extra cost.
Of course there are no tests yet on anything like this, so I request opinions / informed guesses on the following fuzzy question:
==> Will loading the FSB with 4x1GB, or 2x2GB, or even 4x2GB of vanilla unbuffered DDR2 tend to "equalize" the OC potential of these two CPUs? In other words, to limit them both to about the same ceiling? To swamp out any raw clock speed advantage the C1 QX9650 might have?
Or, on the contrary, would making the RAM less "clockable" actually enhance the QX9650's advantage of being unlocked?
Well, it depends what frequecy RAM you choose. Say you go DDR2-1066. If you run 1:1 then go OC crazy and you'll never even OC the RAM, so no big deal. You'd have to hit FSB 533 just to run it at stock speeds (which ain't gonna happen!)
Conversely, if you go DDR2-800, your RAM could limit your OC past 400 FSB on 1:1 ratio.
I have 4 x 1GB DDR2-1066 running slightly overclocked at 1080 right now with some funky ratio. 6:5 i think... (If you want to know exactly, let me know and I'll update it when I get home).