PDA

View Full Version : my overclocking attempts on i7 920 co stepping


funkysnair
08-05-2009, 05:13 AM
well i tried for a while to see if i could get 4ghz, i reached 3.9ghz but was unstable!

i do realise that the co stepping is not the best overclocker next to the do stepping but i did reach a modest 3.8ghz stabble!

i am a overclocking novice so it took me some time, a couple of voltage adjustments (1.26max for cpu 1.65 max for ram)

i wasnt going to go any further than that for sure!

i have now dropped it down to 3.6ghz and benched it some pics.... (1 gpu xfx4890)

i also overclocked the gpu-GPU CLOCK 960mhz MEMORY CLOCK 1100mhz

http://i108.photobucket.com/albums/n37/funkysnair/38stable.png

http://i108.photobucket.com/albums/n37/funkysnair/3dmark.png



if you guys have anymore tips or anything let me know thanx

ZeniTH
08-05-2009, 05:40 AM
crank those voltages!!!!!

srly... we need more info to help you out:

1- description of you hardware
2- bios template
etc...

funkysnair
08-05-2009, 05:47 AM
well im not really after a walk through as ive kinda got the hang of it its just the voltages im not sure about!

intel i7 920 co step
foxconn flaming blade
6gig ddr3 1333mhz g.skill (voltage range 1.5-1.6V) 9-9-9-24
xfx 4890
coolermaster 700 watt 50amp 12volt rail
74gig raptor 16mb cach

good old liquid cooling with ddc pump xspc top tripple rad and ek supreme lt

basicly what can i push max in volts to the cpu and ram?

Worthy
08-05-2009, 05:54 AM
C0, D0, doesnt matter from what I've seen.

Some batches are just better than others with the D0 having chips that are more better than not. In fact some of the d0 chips will clock to a certain point with extremely low voltages and then hit a wall.

My c0 as far as I have attempted did 4.2 with just a modest voltage increase, and does 4.0 by just kicking up the bclk to 200. My batch was 3841A529.

Besides, 3.8 stable on ANY i7 is a force to be reckoned with.

Provide the info that zenith requested and we may be able to break the rut you're in.
It's not necessarily all about the core volts. QPI volts/clocks and even the board used can play a huge role.

funkysnair
08-05-2009, 05:57 AM
thanx for your time lads, always gratefull...

hope theres not too many photos lol, its easyer posting pics than writing it all down



ON ENTERING BIOS
http://i108.photobucket.com/albums/n37/funkysnair/05082009564.jpg

QUANTUM BIOS
http://i108.photobucket.com/albums/n37/funkysnair/05082009566.jpg

CPU FEATURES
http://i108.photobucket.com/albums/n37/funkysnair/05082009567.jpg

MEMORY SETTINGS
http://i108.photobucket.com/albums/n37/funkysnair/05082009568.jpg

ALL VOLTAGE SETTINGS
http://i108.photobucket.com/albums/n37/funkysnair/05082009569.jpg
/\------------- continued -----------\/
http://i108.photobucket.com/albums/n37/funkysnair/05082009570.jpg

OC GEAR
http://i108.photobucket.com/albums/n37/funkysnair/05082009571.jpg

Worthy
08-05-2009, 12:17 PM
Ahhh Phoenix BIOS.

Havent played with one of those in a while. I think since my last MSI board back in the kt333 days LMAO.

for starters (and let me make sure I understand this correctly) CPU turbo on the second shot will run one core at higher multi, correct? If I am correct disable that. It might be the hangup, although I am not positive on that.

Try disabling that, and then kick the bclk up a tad higher and stability check it.
Let me know what happens.


Can anyone tell me what the "system turbo" option is???

Hondacity
08-05-2009, 12:35 PM
you can go as high as 1.375 on your vcore.. its in the box :D

Grnfinger
08-05-2009, 03:48 PM
CPU clock skew will help a lot, ther is no formula for this, its trial and error but its worth sitting down and playing with it.

CPU Vdroop personally should be enabled, I always run it enabled I'm sure others will disagree but its the internet opinions are like assholes everyone has one and they all stink.

Worthy
08-05-2009, 05:06 PM
CPU clock skew will help a lot, ther is no formula for this, its trial and error but its worth sitting down and playing with it.

CPU Vdroop personally should be enabled, I always run it enabled I'm sure others will disagree but its the internet opinions are like assholes everyone has one and they all stink.

Yeah, on my r2e It's called "loadline calibration".
It caused my chip to run at 80c under full load which is why I typically dont recommend using it until a max speed without it is achieved.

I have gotten to 4200 that I have tried without it.
But every board is different.

Grnfinger
08-05-2009, 06:37 PM
on my Maximus II Formula I run it enabled same with my Rampage and Maximus X38 board, it keeps the CPU at a constant vcore and helps stability a lot, I usually run high FSB as I like the bandwidth increase and I'm a bencher so numbers matter...
I'm not saying it's for everyone as I have never used a Foxconn board but I would recommend trying it for the higher clock he may be surprised at the results

funkysnair
08-06-2009, 12:07 AM
cool, thanx for the info...

gonna have another round of overclocking soon

Worthy
08-06-2009, 04:50 AM
on my Maximus II Formula I run it enabled same with my Rampage and Maximus X38 board, it keeps the CPU at a constant vcore and helps stability a lot, I usually run high FSB as I like the bandwidth increase and I'm a bencher so numbers matter...
I'm not saying it's for everyone as I have never used a Foxconn board but I would recommend trying it for the higher clock he may be surprised at the results

I know how you like to overclock lol. On XS there was a maximus formula thread where I played around with one of your templates. ;)

I always kept it enabled on my previous ASUS boards too, until I got this board. I think because of all the power phases it's not needed. All I got was extra voltage and heat. The vdroop on this board isnt bad at all like those might also be why. Who knows?

Read my mo ra review in the reviews section. It's all documented in there. I noticed it when I was testing that rad, a heatkiller 3.0 and was hitting 80c. Took me a while to figure out, but I have the temp data and everything posted in there along with the dialogue leading up to it.

It shows you what I ran into. I was beginning to think the rad or block, maybe even the chip was shit. But it was loadline calibration.

Funky, keep experimenting and see where you get.
We'll be around.

Septim
08-06-2009, 05:22 AM
dang, may need to do this also soon...