Q9450 Overclock

NukeULater

Gawd
Joined
Sep 12, 2006
Messages
917
I just don't understand this. My Q9450 will run 7.5M X 467FSB no problem, it is 11+ hours prime95 stable. But if I use 8M X 450FSB it will crash within ten minutes; Even with much higher voltages. (1.4Vcore, 1.4VTT voltage). Does anyone know how I can hit 3.6Ghz stable without having the raise the voltage to far? http://valid.canardpc.com/show_oc.php?id=579075

My current setup. Rigs specs are in my sig.

3508Mhz_467FSB.jpg
 
Last edited:
what about the voltage you are using on your ram and have you tried increasing the voltage on your northbridge?
 
what about the voltage you are using on your ram and have you tried increasing the voltage on your northbridge?
Yes, I have messed with the other voltages. The northbridge is getting 1.45V, and the RAM is getting 2V and is running under its rated speed.

Thanks for your help. :)
 
Did you try the FSB/VTT voltage? Though your first OC is off a higher FSB, so maybe that's just your processor's limit... Have you tried anything in between? e.g. 445x8 Oh and have you tried using the BIOS rather than Abit's tool? Ya never know...
 
Did you try the FSB/VTT voltage? Though your first OC is off a higher FSB, so maybe that's just your processor's limit... Have you tried anything in between? e.g. 445x8 Oh and have you tried using the BIOS rather than Abit's tool? Ya never know...
Yes I have tried the FSB/VTT voltage. I helps a bit at higher speeds but not at 3.6GHz. Yes all of my overclocking has been done in the BIOS, I was just using the abit tool to show the voltage settings.

I tried 445 X 8, it failed prime95 almost instantly. Yet at 475 X 7.5 it runs prime95 perfectly. I can also get 467 X 8 (3736Mhz) running but it is less stable than 8 X 450.

Thanks for your help. :)
 
Hrm, the fact that it fails 445x8 but does 475x7.5 w/o fault is weird... Are you sure it isn't taking over the memory ratio on it's own or something when you change multipliers? Nothing else comes to mind right now... Sorry.
 
Hrm, the fact that it fails 445x8 but does 475x7.5 w/o fault is weird... Are you sure it isn't taking over the memory ratio on it's own or something when you change multipliers? Nothing else comes to mind right now... Sorry.
Nope, I have the RAM running at 1:1. So currently it is running under it's rated speed and on the stock 2.00V. I have bumped that voltage up, but it didn't have any effect.

Thanks for the help. :)
 
You could try a lower divider nonetheless, maybe the memory controller just doesn't like running it at 950MHz when the rest of the chip is running so far out of spec.
 
3.6 is harder than you'd think. (Or at least harder than I thought it would be, especially in my case having an ES.)
My ES Q9450 can post 4ghz but I couldn't get 3.7 or 3.8 stable with voltage I was comfortable with.
Ended up settling for 3.6, 450x8, ram 1:2 (4 sticks of 800 G.Skill @ 900, 1.9vdimm)
Vcc + Vtt are 1.385, NB 1.3 (IIRC)
Prime95 12hr stable

I would imagine with yours being a regular (non ES) you may have more difficulty than me getting 3.6 without going 1.4+. You might have to go 1.425 or 1.45 to get 3.6 stable. :(
You could probably get away with running 1.45 Vcc 24/7 but definitely not Vtt; I would not go above 1.4 Vtt 24/7 use, not after AT fried their QX9650 above 1.4 Vtt.
 
3.6 is harder than you'd think. (Or at least harder than I thought it would be, especially in my case having an ES.)
My ES Q9450 can post 4ghz but I couldn't get 3.7 or 3.8 stable with voltage I was comfortable with.
Ended up settling for 3.6, 450x8, ram 1:2 (4 sticks of 800 G.Skill @ 900, 1.9vdimm)
Vcc + Vtt are 1.385, NB 1.3 (IIRC)
Prime95 12hr stable

I would imagine with yours being a regular (non ES) you may have more difficulty than me getting 3.6 without going 1.4+. You might have to go 1.425 or 1.45 to get 3.6 stable. :(
You could probably get away with running 1.45 Vcc 24/7 but definitely not Vtt; I would not go above 1.4 Vtt 24/7 use, not after AT fried their QX9650 above 1.4 Vtt.
Well thank you for that information. I did not know about the ES version of this CPU.

One more quick question, at 3.5GHz I was 12+ hours prime stable. Now I can't even play CoD4 without it BSODing. Note, nothing in the BIOS has changed, it is still set to 3.5GHz 467FSB 7.5M. Why would that happen?
 
One more quick question, at 3.5GHz I was 12+ hours prime stable. Now I can't even play CoD4 without it BSODing. Note, nothing in the BIOS has changed, it is still set to 3.5GHz 467FSB 7.5M. Why would that happen?
Okay scrap my last statement. It have done a few more prime95 tests, it was 14.5 hours prime stable. Yet when I play any game it will crash within 5 minutes. I'm starting to think that the video card driver might be corrupted. Anyother ideas?
 
Back
Top