i7 930 @1.2v (4.2Ghz)

M

Milamber

Weaksauce
Joined
Dec 10, 2007
Messages
104
I have had my CPU for around 3 mths now and I just started to test it out. I am able to get the following - I take it this is good volts for the clock?



5F39ofG


DmygBw7

CPU-Z Validator 3.1

It is rock stable as well, did 3dmark11 looping to test.

What kinda clocks do you guys get and volts?
 
Run LinX for 20 passes with full memory usage. Report results.
1.28v or more is need for 4.2 GHz in most cases.
 
I'm on 1.24V for 4Ghz atm. It's stable on a lower voltage but I bump it higher for a bit of leeway as my temperatures are fine. At 4.2Ghz however, I need about 1.3V for 24/7 stability. Any lower than that, linx actually runs fine, but when I try to play Bad Company 2 it just shuts down.

As posted, run linx and see if it holds. Also, you're at 1.25V atm?
 
Also validated the cpuz

EDIT: I forgot you asked for a 20 pass!

sb1chTf

It seems that the volts jump from 1.25v to 1.28 when under load.
 
Congrats you got yourself a nice chip :). Better than mine at any rate, though I'm pretty happy with how mine OCs! Are you on air or water?
 
joxang said:
Congrats you got yourself a nice chip :). Better than mine at any rate, though I'm pretty happy with how mine OCs! Are you on air or water?
Click to expand...

Hi joxang

I'm on air, my whole system is air cooled... since my waterblock last year had a very bad accident.
 
First of all, 20 passes of Linx is usually only adequate if you have a good amount of ram (ie, 4-6GB). Try OCCT and run the linpack test for an hour and see if you're stable. Also, you should be getting faster results--I'd guess that it's due to your small amount of ram...at 4.2 I'm usually getting in the 53GF range.
 
He has 6GB RAM installed, but he didn't click 'All' when running Linx.

Yeh...you have to click the 'all' button next to the RAM. Missed that!
 
mtb2005 said:
First of all, 20 passes of Linx is usually only adequate if you have a good amount of ram (ie, 4-6GB). Try OCCT and run the linpack test for an hour and see if you're stable. Also, you should be getting faster results--I'd guess that it's due to your small amount of ram...at 4.2 I'm usually getting in the 53GF range.
Click to expand...

Eh... what? :confused:

What gave you the impression that I had less than 4Gb of RAM on a i7 rig with DDR3 :eek:

Haha!! I have 6Gb RAM, please check the screenshots, the information is in there ;)

joxang said:
He has 6GB RAM installed, but he didn't click 'All' when running Linx.

Yeh...you have to click the 'all' button next to the RAM. Missed that!
Click to expand...

Ahh bummer, I'm at work now too. I shall click the "all' button and test when I get home.
 
20 runs is also not always stable. Prime on blend for at least 24 hours and some folding will tell you how stable it is.
 
Milamber said:
I heard it was 72hrs for a true reading? 24hrs is just cutting the edge of the cake...
Click to expand...

Why stop at 72 hours? There are another 8693 hours that he could be "unstable".

My point is that you'll never be fully stable. 8-24 hours Prime95 or 20-25 passes with LinX/IBT using all memory should be enough.
 
If you AREN'T using it for folding or using your CPU for heavily threaded applications then 24 hours is enough. So if you are just using it to game, then 24 hours is more than enough. Hell, I would say just do 20 passes on LinX and run it 8 hours or so and you will be good if gaming is all you expect to do.

If you ARE, then I would still recommend doing a short (8 hours or more, preferably 12~24 hours) P95 run + 20~50 passes of LinX , THEN Fold some WUs for a couple of days. After that, you can be decently sure of stability.

Different people have different standards though. Some people want to be absolutely sure of stability and spend a long time checking. Some are satisfied with shorter runs.
 
Well so far, I have played a total of 3hrs Bad Company 2 and 4 hours of Supreme Commander 2 with no issues, ran 3dmark 11 countless times also. I might see how another overclock goes tonight, maybe I will try 4.5ghz...
 
Milamber said:
Also validated the cpuz

EDIT: I forgot you asked for a 20 pass!

sb1chTf

It seems that the volts jump from 1.25v to 1.28 when under load.
Click to expand...

Use a larger problem size, atleast 25,000. 1000 is so low you might as well not have ran the test;).
 
Thread title was a bit misleading. Its 4.2 @ 1.25/1.28 not 1.2 :eek:

But still, nice OC and congrats on the chip ;)
 
Since you're using a Gigabyte board, try messing with LLC (Option #1) and Dynamic VCore.
 
You must log in or register to reply here.
Back
Top