OK, did that: -
FSB @ 140,
vDIMM @ 2.70v,
vCore @ 1.85, "fastest" DIMM settings.
Seemed stable enough, but isn't the slower
RAM clock (140 * 2) an
awful lot slower than default 166 * 2 = 333? Actually, it's showing 350 with FSB @ 140 - as I'd hoped / expected.
I upped the FSB to a measly 141, and Windows wouldn't have it - BSoD.
For now, I'm back to asynchronous
RAM, and calculating Pi to 1m places. More precision than that and it gets unstable "non convergent at SQR05" but since I want to
play rather than draw very precise circles...
*#@%, Hang on... :nervous:
...does the FPU /
CPU get this much hammering when polygoning?
[a new verb] ?
Either way, is "Turbo" or "Fastest" synch @ 280 gonna beat "Fastest" asynch @ 350 ?