Q9550 3.4GHz

Soldato
Joined
2 May 2004
Posts
19,950
Firstly, my specs:
Q9550
P35-DS4 Rev. 1
4x 2GB OCZ Platinum

I'm attempting to get my Q9550 E0 to 3.4GHz.

When I put the bus to 400x8.5 the PC boots into Windows etc. However after a couple of minutes the Windows cursor will stop moving, but I still have the control (i.e. I have to control it blindly). Everything works on the PC, programs start up etc. it's just the mouse that goes a bit strange. I haven't tried Prime while it's running at 3.4GHz, but it'd probably fail.

My best guess is my NB voltage is too low. The 'maximum' supported FSB for this motherboard is 1333MHz, I need it to be running at 1600MHz.

Any suggestions on what I should try first? How much should I up the NB voltage by?

Thanks,
Craig.
 
Try +0.1-0.2V to begin with, see if it's stable, if not, you might want to up something else, though your guess is probably right if you have 4 sticks of RAM in there? Have you upped it before when you put 4 sticks in?

You can have a play as long as you're in safe voltages for the board, though you may have to back down the OC to try and get it stable at stock voltages to a certain point, then up things slowly. It's a lengthly process but one you may have to go through if 3.4 won't be stable. Can't see the E0 not making that pretty easily though. Make sure the RAM timings are absolutely correct too, sometimes some of the later ones can hold back the OC.
 
@alexhull24
Yep, I have 4x2GB 5-5-5-18 8500 1066MHz RAM in the board which is probably putting a fair bit of strain on. I'll put 0.1v on the NB and see if that fixes things. The only voltage I have touched is +0.3v on the RAM itself to get it up to the recommended 2.1v. I have also manually set it to 5-5-5-18.

@OzZie
I thought the default voltage for the NB was 1.2-1.4v anyway? The overclocking section on my board works by putting additional voltage on (i.e. +0.100v), it doesn't actually give the current value.
 
Last edited:
So far, so good. +0.1 on the NB seems to have stabilised at least Windows (I'm not getting the weird mouse thing anymore).

Should I now be lowering the NB voltage to +0.050 or even lower to find the lowest stable voltages I can use?

Cheers,
Craig.
 
First thing I'd do now is stress test it. There'll be no point lowering voltages if it's not fully stable with these settings.

I assume you know how to use prime95 and intel burn test, they're pretty good in finding instabilities. Keep an eye on the temps too.
 
First thing I'd do now is stress test it. There'll be no point lowering voltages if it's not fully stable with these settings.

I assume you know how to use prime95 and intel burn test, they're pretty good in finding instabilities. Keep an eye on the temps too.

Yep, I've used Prime & burn test extensively before :)

I'll run them tonight, if all is well then I'll try lowering the NB voltages.

At idle the CPU is running at pretty much the same as it was with stock clocks, same goes for the motherboard. I'll have to monitor them under load though.
 
Leave the NB and crank up the cpu voltage a tad. I never touched the NB when I oc my e8400 from 3.0 to 3.6, I just upped the cpu voltage, and then slowly droped the voltage agan, until it was unstable and then upped the voltage by 2 notches.
 
Last edited:
Yep, I've used Prime & burn test extensively before :)

I'll run them tonight, if all is well then I'll try lowering the NB voltages.

At idle the CPU is running at pretty much the same as it was with stock clocks, same goes for the motherboard. I'll have to monitor them under load though.
OCCT. It doesn't generate as much heat, but it shows you your core voltages and temps and also does some simple stress. It also has Linpack, which does generate a fair amount of heat.

My Q9550 VCore was actually set by the BIOS to be 1.25000V on Auto, OCCT reports it as 1.3 though, so I'd imagine it fluctuates for whatever reason on that setting.
 
While I can't remember any of the settings for the Q9550 anymore, I would like to add that 3.4Ghz on a Q9550 should be easy peasy. It is a great chip that can be pushed quite a lot harder. 3.4Ghz is a good overclock though :)

What cooler do you have? What temps are you getting?
 
3.4 should be do-able at stock volts on an E0. Mine is at 3.8 rock solid stable, although it does need 1.32v.
 
While I can't remember any of the settings for the Q9550 anymore, I would like to add that 3.4Ghz on a Q9550 should be easy peasy. It is a great chip that can be pushed quite a lot harder. 3.4Ghz is a good overclock though :)

What cooler do you have? What temps are you getting?

65C is maximum after quite a while under Prime, I've got the Scythe Ninja.

3.4 should be do-able at stock volts on an E0. Mine is at 3.8 rock solid stable, although it does need 1.32v.

The CPU is at stock, the only reason I had to up the NB voltage slightly is because I have 4x2GB 8500 1066MHz RAM & the motherboard officially only supports 1333MHz FSB.

CPU voltage is 1.18v according to Everest, VID is 1.2125v according to the BIOS & Core Temp.

My PC is now running stable at the current settings (3.4GHz CPU, +0.1v on the NB). I think I'll have a play around with the NB voltages and see how it runs at +0.050.

Leave the NB and crank up the cpu voltage a tad. I never touched the NB when I oc my e8400 from 3.0 to 3.6, I just upped the cpu voltage, and then slowly droped the voltage agan, until it was unstable and then upped the voltage by 2 notches.

The CPU isn't limiting my overclock at all. From what I've read and seen it can easily do 3.4GHz at stock voltages. I now know for sure it's my motherboard limiting this overclock as I have to add +0.1v to the NB to stabilise things.
 
Last edited:
The CPU is at stock, the only reason I had to up the NB voltage slightly is because I have 4x2GB 8500 1066MHz RAM & the motherboard officially only supports 1333MHz FSB.

CPU voltage is 1.18v according to Everest, VID is 1.2125v according to the BIOS & Core Temp.

Ahh sorry, that'll teach me to skip-read :o
 
Back
Top Bottom