It may just be the oem 9650 that I got, but it doesn't appear to me that there is a huge difference between good air and single phase in terms of stable overclocking of these. I still haven't found the exact top end of mine here, but it seems like it is somewhere in the 4800mhz area. And that needs 1.55 volts. (this is on a p5k dlx and ddr2 setup) That's not much higher than some of these air results. Actually this was somewhat similar in the QX6850 to some extent from my experience. I would have expected alot more, but I guess when your overclock a cpu from 3 ghz to 4.4 on air, that's already close to the limit of the cpu short of liq nitrogen, or multi stage phase cooling. Not that there is anything to complain about 4800, but with those super high air results, it seems like there would be more headroom. Again, maybe it's just my part. I did have it posting in windows at 49xx on a P5E3 board with 4GB ddr3 before the MB went belly up, but the phase evap head mount wasn't making full contact and it wasn't stable.
I also wanted to mention that high fsb on this 45nm one seems much harder to achive than on my QX6850. I see another user reporting the same here:
http://www.xtremesystems.org/forums/...d.php?t=166087
Maybe the extra cache or something. Not sure why.
Reply With Quote
Bookmarks