Retail Q9650 Overclocking Thread

[NapalmV5]

Well I use those exact voltages for my 4050 450FSB 24/7 stable in prime95, and 50 runs of linpack. I'm actually typing from it now, even. And I don't use bios or software for voltages, ever, DMM is your friend. The problem is CPU-Z.

Admittedly, I have QX9650, C0, But I can guarantee that you need some tweaking, or it's teh cpu, not the board, as this board has provided more FSB than any other I have tried so far, ever, for 45nm quad.

I can easily boot into 460FSB, but I have not done much work on stability yet(this is P05 bios):

do you have CPU GTLREF and MCH GTLREF settings in the bios on your new board?..if so, work with those.

guys its not my cpu/tweaking.. its not my first GTL settings mobo..

right now its in even worse shape.. cant even boot @ os @ 3600 450x8 only 400x9

why does the cpu run rock stable @ 4050 450x9 on the dfi dk x48t2rsb+ ? albeit @ .05 higher vcore than on mars but without any GTL tweaking

anyways.. ill be getting another blackops.. hope not all blackops are the same as this one

[cadaveca]

why does the cpu run rock stable @ 4050 450x9 on the dfi dk x48t2rsb+ ? albeit @ .05 higher vcore than on mars but without any GTL tweaking

'Tis more than possible that auto values in DFI bios are higher than blackops, and actually, that's what I'd point a finger at(i think Ace-A_Rue agrees with me). This is why you must never overclock on auto settings...how can you know what's wrong, if you do not know what's set?

This is not a slant directed at you, but at mobo makers. The whole label of "AUTO" should no be so...it creates more confusion than is needed! If you knew GTL values of both boards, and that they were the same(personal experience from overclocking the boards myself tells me that they are not), maybe you could sort the issue.

Read-to-write delay is also quite tight on this board, and affects ram clocks, so much so that my ram is NOT stable @ jedec timings, because the board ignores some SPD info.