E6600 Overclocking Help

Buckster

Buckster

Buckster

Soldato
Joined
20 Aug 2006
Posts
9,778
I'm very pleased with my E6600 -its performing well, but of course its always nice to push it as far as it can go (without going stupid)

I have a P5B, Corsair TwinX 6400C5 RAM, E6600 (Week 24)

RAM set to 1:1, 4-4-4-12 @ 1.9V. C1E Disabled, EISA Disabled

At 331fsb the CPU runs great, in fact I can run it a bit under auto-voltage, at 1.34V, and its stable, max core-temps (using core-temp) are 50C.

Above 336 fsb I can't get anywhere, Orthos crashes after about 40seconds-2 mins, so not stable.

I tried Vcore up to 1.4125 in Bios (which is about 1.4 real), and Voltage of Dimms up to 2.0V, and also dropped them to CAS5, no change, still not stable.

have I reached the limit of the CPU within sensible voltages, or is the Voltage on the Northbridge or something holding me back (all Vs apart from CPU and RAM are set to Auto)

Thanks, Mark.
 
Try increasing the FSB voltage (chipset voltage) a little and see if that works. You probably raised your FSB too high and not enough voltage is being supplied. You should be able to get over 340 as long as temps allow.
 
It could be any of the components holding you back. Start again from the beginning and try to establish the 'ceiling' overclock for each component in your system.

Lets break this down a bit. Your CPU can be overclocked by increasing the FSB, and if this causes instability then you may be able to rectify that by increasing the CPU voltage. Your FSB can also be increased, and if this causes instability then you may be able to rectify this by increasing the chipset voltage. Also, your RAM can be overclocked by increasing the FSB, and if this causes instability then you may be able to rectify this by increasing the RAM voltage.

Can you spot the link in the above?

They all involve increasing the FSB, so how are you supposed to know what caused the instability?

Well, the simple answer is you can't. Unless you take measures to eliminate the other components from causing the instability. This is where the multipliers come in

Dividers are just ways of increasing/decreasing the speed of components relative to others, and independently of the FSB. If you decrease the CPU/FSB multiplier, you can eliminate this from being the component that is causing the instability. If you decrease the RAM/FSB multiplier, you can eliminate the RAM from being the component that is causing the instability. If it's still unstable after decreasing these multipliers, then it's likely that the mobo is causing the instability, and voltage can be added to the chipset to try to increase the overclock.

Try this:

1) Set your RAM/FSB multi to around half of what it should be (i.e. 1:2 ratio instead of 1:1);
2) Set your CPU/FSB multi to around half of what it should be (i.e. 4x or 5x, instead of the usual 9x - I think ** CPU should be running 266x9 at stock, right?)
3) Try increasing your FSB in 5 or 10 Mhz increments and stability test between. If you experience instability, increase the chipset/NB voltage and try again. If it's stable, continue to increase the FSB. Keep doing this until you find a stable ceiling. Keep a log of all the FSB's and voltages and whether it was stable or not.
4) You now know what the max FSB is that your mobo can handle. Put the FSB back to stock, set the CPU multiplier back to 9x and start overclocking again. This time, keep your chipset/NB voltage inline with the notes you made for the frequency it's running at, and increase the CPU voltage if you get instability. Keep doing this, increasing in 5 or 10Mhz increments until you find the stable ceiling for the CPU. Keep a log!
5) Set the CPU multi back down to 4x or 5x. Set the RAM multi back to 1:1. Start increasing the FSB, chipset/NB voltages once again in line with the log you made, and increase in 5 or 10Mhz increments until you find the stable ceiling for the RAM.
6) You should now have the max stable overclocks for each individual component, so now you need to try to work all of them together by calculating the frequencies and dividers to try to get the best compromise for them all. Prioritise on the CPU speed first.

I hope that helps :) And btw, I'm an Athlon64 overclocker so if I've made some fundamental mistakes in the nature of overclocking Core2Duo's then please somebody feel free to correct me
 
Pesonally I'd say that you need to slacken your RAM timings. Remember that the SPD for your modules are 5-5-5-12 at 400mhz, so the ICs on these modules would not appear to be that great.

Try them at 5-5-5-12 and see if that gets you any further. Also try upping the vmch and vcore.
 
will give all you've suggested a try - thanks.

Kesnel - the RAM chips are only running at 2x336 = 672 CAS4 - vs rated of 666 CAS4.

I tried it at CAS5 - no difference, just as unstable :(

Mark.
 
the answer for me was to put your memory in slots 1 and 2 ie not in DDR mode...

worked for me on 2 rigs, could not get past 320ish fsb without loads of errors....

i can hit 430fsb with e6300 now on 2 p5b boards, its mem compatibility....bios should fix this
 
U think i should do that as well hardster. move my ram beside each other insteada dual mode. i cant seem to get any higher than 350fsb.
will the ram run slower if its not dual. ??
im running a DS3 and 2gb of crucial ballistix pc8000 1000mhz ram.

c
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom