- Jan 11, 2001
- 28,830
- 17
- 81
I've heard through the rumor mill that the P45 chipset was a PIA to OC with, coming from P35 I found that hard to believe but alas, I took the plunge into P45 anyway and grabbed a MSI P45 Neo3-FR. If you care to see my initial progress and build info, it's cataloged here: http://forums.anandtech.com/me...=2208946&enterthread=y
Using my (now) trusty E8500, I had no problems with the P45 and pushing the FSB north of 500 to get a prime stable OC of 4.25GHz with 1.41v core and only minor tweaks to VTT and NB volts. Everything else was left untouched. Top FSB into windows (not tested prime stable) so far was 545fsb but being I run a set of Buffalo Firestix which are micron D9, I refuse to push the ram voltage past 2.1v so I think my ram crapped out before the mobo. Since I have no other ram, i'll never know what the true top limit is as i'm not willing to sacrifice my firestix for an all out FSB run.
I began to think all those P45 naysayers were full of bunk and the P45 was as easy to OC as the P35 chipset. Well, that brings me to today when I finally got the chance to properly mount my Q9450 and give it a run in the P45. All I can say is this thing is a PICKY sob with a quad in the socket!!! Nudge ANY voltage one bin up or down and it'll post fine but refuses to load windows. I've had it into windows at 3.7GHz but it will only prime for about 30 mins. Sadly, I got too tired of messing with the voltages to keep trying so I backed all voltages AUTO except Vcore at 1.2 and set the fsb to 400 just to get into windows and type this up. When they say overclocking the P45 is a blind art, BELIEVE IT!
With 1.35V core, Dram on 1.9v, MCH on auto, VTT on 1.3V and ICH on auto it'll run and prime fine for about 30 minutes. Touch any of those voltages one notch in either direction and it no longer loads windows at all. I've tried randomly running voltages up and down in an attempt to gain further stability but my efforts are futile. I'm gonna let her run at 3.2 for the time being and allow my frustrations to subside before I try to tweak further. It looks as though my prime stable 3.6GHz quad chip still manages to elude my grasp! :frown:
Using my (now) trusty E8500, I had no problems with the P45 and pushing the FSB north of 500 to get a prime stable OC of 4.25GHz with 1.41v core and only minor tweaks to VTT and NB volts. Everything else was left untouched. Top FSB into windows (not tested prime stable) so far was 545fsb but being I run a set of Buffalo Firestix which are micron D9, I refuse to push the ram voltage past 2.1v so I think my ram crapped out before the mobo. Since I have no other ram, i'll never know what the true top limit is as i'm not willing to sacrifice my firestix for an all out FSB run.
I began to think all those P45 naysayers were full of bunk and the P45 was as easy to OC as the P35 chipset. Well, that brings me to today when I finally got the chance to properly mount my Q9450 and give it a run in the P45. All I can say is this thing is a PICKY sob with a quad in the socket!!! Nudge ANY voltage one bin up or down and it'll post fine but refuses to load windows. I've had it into windows at 3.7GHz but it will only prime for about 30 mins. Sadly, I got too tired of messing with the voltages to keep trying so I backed all voltages AUTO except Vcore at 1.2 and set the fsb to 400 just to get into windows and type this up. When they say overclocking the P45 is a blind art, BELIEVE IT!
With 1.35V core, Dram on 1.9v, MCH on auto, VTT on 1.3V and ICH on auto it'll run and prime fine for about 30 minutes. Touch any of those voltages one notch in either direction and it no longer loads windows at all. I've tried randomly running voltages up and down in an attempt to gain further stability but my efforts are futile. I'm gonna let her run at 3.2 for the time being and allow my frustrations to subside before I try to tweak further. It looks as though my prime stable 3.6GHz quad chip still manages to elude my grasp! :frown: