A lot of us are having problems overclocking the DQ6 with the F8 BIOS that we didn't have with the F7 BIOS.
The machine would fail to POST when raising the FSB
I wired an external pushbutton to clear the CMOS on my DQ6, and started experimenting.
The answer seems to be FSB voltage. I ended up adding added +0.25v to the FSB before I could get it to POST at 400MHz. With the F7 BIOS, the DQ6 only needed +0.10v to run at 400MHz.
So, right now, I'm running a processor multiplier of 7 (2.8GHz), FSB 400, memory multiplier 2. It passes Memtest and seems to be doing OK with Orthos (its running Orthos while I'm writing this). Temperatures are good too.
I know this E6600 will do 3.2GHz easily if I just change the processor multiplier to 8, but I'm going to run it like this for a while.
I've updated my DQ6 to F8 BIOS and got something like yours too. Previously I've been running my cpu at 3.2ghz (400x8) with 1.3v vcore bios, default vmch & vfsb for 24/7. But now, it always automatically sets back to default fsb clock when I booted it. The different is mine could pass 20 minutes orthos at 2.8ghz (400x7) with everything default volts.
And for benchmark test, now I have to add my v.core to 1.525v for 3.6ghz (400x9) :?
However there is one advantage from F8 BIOS. Now I've no problem running my GSkill 6400HZ (Micron D9GMH) at DDR1000 4-4-4-12 2.2v which cannot be done with my last F7 BIOS (it needs 2.3v to orthos stable).
I just flashed my DQ6 to F9 beta bios, guess what? many problems have been solved, cold-boot problem and FSB 400+ problem. Finally, this is the first time ever I reach FSB500 without super-pi error in timing.