stupid question?

I saw in the screenie above you were overclocking the memory, I was expecting to see approx 533MHz (1:2), Is the FSB 266MHz in that shot or 272MHz :confused:

Yeah, didn't realise the FSB was at 272, reduced it down to 266

have followed your instructions and these were the results you said to look out for

CPU - 2.39MHz
vcore idle - 1.300V
vcore load - 1.280V
temp idle - 42C
temp load - 54C

CPU - 3.01MHz
vcore idle - 1.300V
vcore load - 1.260V
temp idle - 42C
temp load - 61C

all went fine using prime95 and games, no instability

Run a 3.0GHz Quad Core on a 333MHz-FSB/1333MHz System Bus with the Memory at DDR2-1066 through a few games, diagnostics etc and come back when your done! :cool:

only problem was when i set the FSB to 333MHz, DDR2-1066 wasn't an available option, i went for the nearest one which was 1002. have i gone wrong somewhere?
 
good results, that's strange you cannot select DDR2-1066 when using a 333MHz-FSB, I was hoping your board had the [5:8] setting :confused:

Just had a quick read of a few reviews and it seems these are the available memory options:

FSB clock x 1.0, 1.2, 1.25, 1.5, 1.6*, 1.66, 2.0

* missing

So with a 333MHz-FSB that would mean running the memory at either of these two (you did the first!)

1.5 = 500MHz (DDR2-1000)
1.66 = 552MHz (DDR2-1105)

It would be interesting to know if your Kingston HyperX can run at DDR2-1105 but that would take a few hours of testing so maybe scratch that for the moment and come back to it later.

I think the extra memory multipliers may be something Intel brought in with the newer P45 Express chipset. Seems strange that they say the board supports 1333MHz-System Bus chips and DDR2-1066 but not let you do both at the same time :confused:

  • Support Intel® Core™ 2 Extreme/Duo/Quad processors with 1333/1066/800MHz FSB
  • 4 X 240-pin DIMM sockets support max. memory capacity 8GB
  • Supports Dual channel DDR2 1066/800/667 Un-buffered Non-ECC memory

Do you have the latest BIOS installed?

Ok so we are nearly there with the meat and potatoes overclock but the next bit could be tricky . . .

With the FSB set to 333MHz go into BIOS and drop the CPU Multi from [x9] to [x7] then Save & Exit from BIOS. If the computer POSTS let it boot to Windows and run a quick test/game to see if everything is ok. Your CPU should be 2.33GHz (7x333), Memory Still 500MHz (DDR2-1000) but behind the scenes your Northbridge chip will be running at about 428MHz (similar to a 428MHz FSB). We did a similar test before when you set 6x266 which you reported worked fine so this is just a little bit harder.

If 7x333 is fine then reboot back into BIOS and change the CPU multi from [x7] to [x6] then Save & Exit from BIOS . . . . there is a slight chance your PC may not POST at this stage, if it does your doing well. If by some chance you can boot at 6x333 let it go to windows and test it. Your Northbridge chip will now be running at 500MHz which is quite a bit above its nominal design and you may have problems . . . . bit of testing/priming/gaming would let you know. Also as your voltages are on [Auto] still keep a beady eye on your Northbridge temps, not sure if you can do this through software but you can certainly wodge your thumb on the NB heatsink taking care that it may be as hot as a iron. . . . hopefully its just warm! :D

6x333 = 2.0GHz CPU
Memory = 500MHz (DDR2-1000)
Northbridge Clock= 500MHz

If this passed we can push on! :cool:
 
Last edited:
haven't updated the bios in quite a while, should i?

right... got it booted at...
6x333GHz
and DDR2-1002

northbridge is warm but not burning :)

didn't die during games/prime95 :)
 
If you open CPU-z and look at the motherboard tab what version BIOS do you have?

The latest version is #17 here (05/26/2008)

Would be nice to have all the latest BIOS code on your board no? :)
 
Sorry to but in on your brilliant walk though (its helping me learn a lot about overclocking) and the thread, but how does lowering the multiplier cause the northbridge to run faster? Also with the volts that nicnac posted above, why is there less volts at load then idle, shouldn't it be the other way?
 
how does lowering the multiplier cause the northbridge to run faster?
The frequency of the Intel® AGTL+ Northbridge which most folks refer to as the FSB is generated by something known as the NBCC (Northbridge Core Clock). The NBCC works out its frequency by performing the following sum:

Native CPU Multi divided by BIOS set CPU Multi = X

X multiplied by BIOS set FSB = NBCC MHz/frequency


As an example lets take nicnac1's Q6600 running at 9x333=3GHz Approx

Native Multi= x9
Set Multi= x9

9/9=1

1 x 333 = 333MHz NBCC

9333.gif


Now when we drop the CPU multi from it's native [x9] to [x6] look at the maths

Native Multi= x9
Set Multi= x6

9/6=1.5

1.5 x 333 = 500MHz NBCC

6333.gif


That's a very sweet 50% increase in bus bandwidth by raising the NBCC from 333MHz to 500MHz! :eek:

As you can see the FSB doesn't change in CPU-z, you can only read the NBCC MHz by reading it in SiSoft Sandra . . .

The only exception to this rule is the Intel® Core™2 Extreme Processor's (the $$$$ ones with unlocked multi) which the NBCC reads as native no matter what multiplier is set.

why is there less volts at load then idle, shouldn't it be the other way?
It's vDroop and most common on older motherboards, also effected by the quality of ones PSU also the amount of PWM's built into the motherboard. Some people can work with it I prefer to have it the other way around i.e the CPU uses less voltage when Idle and more Volts when loaded. ASUS have a feature called LoadLine Callibration (LLC) built into their newer boards that largely negates vDroop.

For peeps who don't have LLC they just need to dial in a greater amount of vCore at a BIOS level to compensate, it's vital the chip gets enough juice when it needs it most otherwise its crash-city. The only downside to this is that the Chip gets a humungus amount of juice supplied to it when idle! :p
 
I flashed the bios...
all went well, installed the new bios...
5 system beeps...
no signal to the monitor but computer is still running...
frightened to reset incase it screws the bios up...
what should i do?
 
*If* the flash process has completed and the computer rebooted but failed to POST its safe to power it down.

Once powered down try to start as normal, if that fails then power down again, disconnect from the mains and find the Clear CMOS Jumper and change the jumper, wait a few seconds then put the jumper back as normal and then try to power up again!

Couple of questions, did you reset your overclock and flash from a stock set-up? and scared to ask but what did you use to flash? :eek:
 
No that sounds fine?

Can you describe the flashing process, did you get a confirmation at the end it was successful or similar?

Got it working yet?
 
no confirmation, just a blank screen then powerdown.
started up the machine and it says it's testing memory and taking an age to do it... can't escape it. the led on the mb is saying C1 initially then 52 if that helps
 
Ok let me just pick myself up from the floor! :D

Reboot back into BIOS and Load Optimised Defaults

Save & exit then straight back into BIOS and have a nose around, see if there are any new options or options missing!

It may be something in a newer BIOS is causing that long memory count (Quick Power On Self-Test = Enabled) or 4GB will take yonks!

If we can sort this then leave the processor at stock and get that memory running at DDR2-1066 again for real quick test!
 
Last edited:
Ok let me just pick myself up from the floor! :D

Hehe - that's the only downside with these kind of threads, you suddenly feel a sense obligation and guilt if things go t*ts up... (i hate waiting for replies to 'bios-flash' advice – even when inserting all the disclaimers.)

PHEW!
 
Yeah I know what you mean! :p

I should have either got the BIOS flash done first thing and out the way or left it for another day, reckoned it was worth doing though, seems like a fair few changes and I was mainly curious to see if it improved the memory options?

ip35probios.jpg
 
Back
Top Bottom