i flashed the bios to 12.B03 and realized it wasn't a good bios at least from the standpoint of allowing me to raise the multi...returned to 12.B02 and that works very well with the QX9650.
i am running wPrime 1024M stability bench to see what errors are induced as i climb the FSB ladder...so far, i am at 3870 mhz using 1.28 vcore set in the bios...it droops to about 1.25 under load (0.03 Vdroop under load)...Uguru's vcore readout is very close to actual vcore using a DMM to read off the board...as to vcore regulation, it is working a lot better than the IP35 Pro...first of all...i literally have no Vdroop for idle conditions...so...i guess the digital PWM has been markedly improved...that is a definite plus in my book.
don't know about temps since nothing is accurate as far as i can tell...Uguru is stuck on 13C for idle up to 3870 mhz...beta coretemp 095.6 is questionable as to accurate temps...cores 1 & 2 are showing 57C while cores 3 & 4 are showing 45C during the 1024M bench.
on a positive note, i'm feeling good with only making vcore increases...haven't touched MCH, CPU VTT, ICHIO and ICH...the only other one is Vmem and that is in auto so it is running around 1.97v (actual) with my GSkill's DDR800 in 1:1 setting at the moment...those are hanging decently with the 1024M bench...not going to toy with the timings until i find the top end of the overclock using this monster Intel heatsink....right now, they are running 5-5-5-15
EDIT...still searching for the upper limit of vcore for 4052 mhz to complete 1024M bench...i increaded the sub-voltages to a small degree with so far no help...i might need to jump a multi so i can lower the FSB down to around 404.
Reply With Quote
Bookmarks