q6600 @ 4ghz on air

[alexp999]

What if you dont have a box?

Can you find out of software?

Nope, it aint programmed into the chip, so cant be read from it.

Its only on the sticker on the box, or on the IHS of the chip.

 

[Sukh]

Thanks, I will have to have a look for the box!
Just curious really as mine gets to 3.6 at 1.3volts so was just wondering if it could make 3.8-4.0

 

[alex24]

I also share the opinion this isn't really possible on air. On water, yes, though it'd still need a decent system as past 3.6 they eat up the volts!

I've got a L740A batch one with a 1.2125 VID that clocks very easily (currently for sale on MM -sold pending payment atm tho). I'm sure under water it could reach 4. How stable it would be is debatable. Depends how you define stable!

 

[setter]

Most i ever eeked from mine was 3.8ghz, ran it most of the time at 3.6ghz, 1.400 vcore, chip had a vid of 1.2750, to hit 3.8ghz i had to go slightly over 1.5 vcore, lap the cpu and my TRUE and use a pair of 2000 rpm fans on the cooler.

8hr prime screens

3.8ghz

3.6ghz

Vcore for 3.6ghz was initially just over 1.4, managed to get it down to 1.400 and maintain stability, however the jump required to go higher was quite signifigant.

 

[Greebo]

Pretty sure my q6600 under water needed 1.6v for 4ghz prime stable.

That's a lot of voltage and hence heat for air cooling to handle IMO

 

[Sukh]

Is the voltage needed for the extra mhz non-linear?
E.g from 3.4 --> 3.6 needed and extra 0.1
But 3.6 --> 3.8 need 0.2 or something?

 

[setter]

With mine, i ran it at 3.0ghz on stock volts of 1.2750, then jumped to 3.4ghz on 1.3750, initially tested it at 1.41 ono for 3.6ghz, later got this down to 1.400, 1.50625 in bios for 3.8ghz, 1.512 at load, llc enabled on an asus p5q deluxe.

 

[Sukh]

Whats LLC? (sorry)

 

[alexp999]

Load line calibration, compensates for vdroop

 

[alex24]

It's definitely non-linear. In fact, it'd be interesting to see how the stable voltage varies with clock speed, and see if there are patterns for different cpus, VIDs or batches. Maybe then you could predict how well yours would clock!

 

[ultrakipper]

Will your memory cope with an overclock ? , I have normal 800mhz memory and its at 1:1 at 3.6

 

[Pneumonic]

Is the voltage needed for the extra mhz non-linear?
E.g from 3.4 --> 3.6 needed and extra 0.1
But 3.6 --> 3.8 need 0.2 or something?

Graph shows the average increase in vcore over VID required for a clock:

 

[Pneumonic]

For reference my old Q6600 would do:

3.6 ghz - 1.32v
3.7 ghz - 1.40v
3.8 ghz - 1.45v
3.9 ghz - 1.49v
4.0 ghz - 1.53v

 

[Sukh]

Sorry, I'm sure I asked this before, is that the BIOS volts with/without LLC or the acutal volts shown by CPU-z?

If so, mine does 3.6 at 1.29v (cpu-z) now so would hopefully make 3.8!

Also thanks for the graph, are the spikes due to that being some peoples max OC and having to up the volts, skewing the population of data?

 

[Pneumonic]

yes the graph is a combination of overclocks from this forum and XS so in most instances are averages for each clock and more than likely maximum overclock's.

All voltages are load voltages (under Prime95 small ffts) as shown by CPU-Z

 

[Sukh]

So roughly, vid 1.252 + 0.25 = 1.5v to get to 3.8ghz, think I shall stay at 3.6/1.29v