Road To Overclocking

87dtna said:
LOL.

OK well I never tried the 2209 bios, so that must be it.

I always disabled the SAS controller in the bios. It cuts about 15 seconds off the boot up time as well, that thing is soooo annoying. I forget what tab it's in though. Possibly boot tab, but it might be in advanced in a submenu. It's like hardware bootup or something like that, you'll see the SAS controller in there just disable it and you're good.

Yeah I've owned a gazzilion cpu's and also video cards. So those are really mostly the two I spend time in. Motherboards every now and then, but there again I don't remember the bios layouts of all those boards I've had so sometimes I can't help.
Click to expand...

Will turn of the sas controller.Bloody annoying marvell sas controller taking soo long to boot.
I always hated that.
 
OK so anyway back the the OC.

You say that 183x21 at 1.31v is stable? But 200x19 is not even at 1.325v? And 200x17 is?

Makes no sense whatseover....LOL. Thats why when I remembered the turbo cut I was like Ahhhh thats it! It's really only running at 3.6ghz under load. huh.

Try 200x18 and see if that is stable at 1.31v.
 
87dtna said:
OK so anyway back the the OC.

You say that 183x21 at 1.31v is stable? But 200x19 is not even at 1.325v? And 200x17 is?

Makes no sense whatseover....LOL. Thats why when I remembered the turbo cut I was like Ahhhh thats it! It's really only running at 3.6ghz under load. huh.

Try 200x18 and see if that is stable at 1.31v.
Click to expand...

The 200x17 is stable but the 200x19 is not stable even at 1.325V.
And even the 200x18 is stable.

Are you sure about the turbo cut on my motherboard since I have never seen my multiplier go below 21 and since day before yesterday onwards my CPUz is continuosly on.
 
No I'm not sure because I never used the new bios.

Try 1.3625 just for the heck of it at 200x19. If the extra Vcore makes it stable, then the turbo multi HAS to be cutting.
 
87dtna said:
No I'm not sure because I never used the new bios.

Try 1.3625 just for the heck of it at 200x19. If the extra Vcore makes it stable, then the turbo multi HAS to be cutting.
Click to expand...

I tried till 1.4 but it is still unstable.200x19.
Is there something wrong with my motherboard or the temps are too high or the chip??
 
Try Small FFT prime95 instead of blend. I'm curious. LMK what your temps get up too as well.

Also, set QPI voltage to 1.35. Your ram is still at 1200 right?
 
87dtna said:
Try Small FFT prime95 instead of blend. I'm curious. LMK what your temps get up too as well.

Also, set QPI voltage to 1.35. Your ram is still at 1200 right?
Click to expand...

I did all stability testing with small ffts only.Temps go upto 87C.
And my RAM is 1200 only.Then there is the BSOD.After 15mins or so..
 
OK, temps are fine (for smallFFT).

Did the extra volts prolong the stability at all? Or did they make no difference and you'd BSOD around 15 minutes anyway?

Did you try 1.35v QPI voltage?

Also, XMP off now...no difference?
 
mihir said:
I did all stability testing with small ffts only.Temps go upto 87C.
And my RAM is 1200 only.Then there is the BSOD.After 15mins or so..
Click to expand...
Those temps are still scaring me a bit. It's definitely caused by your high ambient temperature, but unless you can get the ambient temperature down, you might want to get a liquid cooling system. I've never tried liquid cooling, so it's probably good to see what some other members think about it first. ;)
 
87c is fine for Prime95 smallFFT. Prime95 will stress the CPU far more than any real life app could possibly do, he'll never see over 70c doing anything normal if Prime95 smallFFT is getting to 87c. You also have to take into account smallFFT heats about 5c more than blend prime95.
 
87dtna said:
OK, temps are fine (for smallFFT).

Did the extra volts prolong the stability at all? Or did they make no difference and you'd BSOD around 15 minutes anyway?

Did you try 1.35v QPI voltage?

Also, XMP off now...no difference?
Click to expand...

yes the extra volts prolonged the stability only a bit like +1/2 mintutes after that it all froze.
Yes with XMP turned off and 1.35V QPI.


@spynoodle - The temps scare me too,and I cannot afford a liquid cooling system since I will have to get everything imported from USA.I can maybe buy it part by part but I would rather not spend so much.
 
mihir said:
yes the extra volts prolonged the stability only a bit like +1/2 mintutes after that it all froze.
Yes with XMP turned off and 1.35V QPI.


@spynoodle - The temps scare me too,and I cannot afford a liquid cooling system since I will have to get everything imported from USA.I can maybe buy it part by part but I would rather not spend so much.
Click to expand...
Yeah, they're ridiculously expensive. That's why I don't have one. :o I guess that you could just try to get as much air over the thing as possible. However, as-is, those temps really aren't safe.

If you want to air on the adventurous side (haha, "air." It's punny. :)), you could try to to put a Peltier plate on your CPU:
http://en.wikipedia.org/wiki/Thermoelectric_effect
^That's one of my current projects. I modded my RVT-12025D to include thermoelectric plates, but I still can't get it to cool right. :angry: I think that there are also some condensation measures that you have to take, but I'm not 100% sure. It definitely takes some work, but it's worth a shot if nothing else helps. ;)
 
Last edited:
Now I have the True in a push pull and some better airflow.
The Push fan will pull air from outside the case where my air cooler is situated.
and the pull will pull the hot air from the HS and push it towards the back of the case where another exhaust is place which pushes the hot air outside.
So I will be trying for 4Ghz.

I want to know what Memory timings should I keep and what memory clock. :D
thanks


FIRST POST

So I finally decided to overclock even thought I have a hot-chip but I just decided to F@#K that.
NOTE:This is the First Time I overclocked a CPU.

So I used two guides
First this one - http://www.bit-tech.net/hardware/cpus/2008/11/06/overclocking-intel-core-i7-920/1

Which was a quick and dirty and a brainless overclock.

And another guide which I quite liked - http://www.overclock.net/intel-cpus/538439-guide-overclocking-core-i7-920-930-a.html

For all methods my turbomode,speed step and CiE and all spread spectrums wre turned off and LLC was turned On.
Using the first method.

My setting were
BCLK Frequency - 200(MHz)
DRAM Frequency - 1,600MHz
CPU PLL Voltage - 1.96V
QPI/DRAM Voltage - 1.35V
DRAM Voltage - 1.66V


Which gave me an unstable overclock of 4.0GHz+.I only had enough time to run Superpi mod and got the score of 10.28s for 1M.


Then since i encountered BSOD 0x124.

So I then I tried the other guide.
So here were my settings.

Multiplier - 21x(For all)


First Try.
BCLK - 143MHz
Memory - 8x BCLK
QPI - 18x BCLK
Uncore - 2x BCLK
CPU PLL - 1.88V
VCORE - 1.3V
QPI/Uncore - 1.3V

Got a stable overclock of 3.003GHz


Second Try
BCLK - 163MHz
Memory - 8x BCLK
QPI - 18x BCLK
Uncore - 2x BCLK
CPU PLL - 1.88V
VCORE - 1.3V
QPI/Uncore - 1.3V

Got a stable overclock of 3.423GHz

Third Try
BCLK - 183MHz
Memory - 8x BCLK
QPI - 18x BCLK
Uncore - 2x BCLK
CPU PLL - 1.88V
VCORE - 1.3V
QPI/Uncore - 1.3V

Unstable Overclock of 3.843GHz

Fourth Try
BCLK - 183MHz
Memory - 8x BCLK
QPI - 18x BCLK
Uncore - 2x BCLK
CPU PLL - 1.88V
VCORE - 1.3125V
QPI/Uncore - 1.3125V
Stable Overclock of 3.843GHz


Fifth Try
BCLK - 203MHz
Memory - 8x BCLK
QPI - 18x BCLK
Uncore - 2x BCLK
CPU PLL - 1.88V
VCORE - 1.3125V
QPI/Uncore - 1.3125V

Failed.


Sixth Try
BCLK - 203MHz
Memory - 8x BCLK
QPI - 18x BCLK
Uncore - 2x BCLK
CPU PLL - 1.88V
VCORE - 1.325V
QPI/Uncore - 1.325V

Failed.


Seventh Try
BCLK - 203MHz
Memory - 8x BCLK
QPI - 18x BCLK
Uncore - 2x BCLK
CPU PLL - 1.88V
VCORE - 1.3375V
QPI/Uncore - 1.3375V

Failed.



Eightth Try
BCLK - 203MHz
Memory - 8x BCLK
QPI - 18x BCLK
Uncore - 2x BCLK
CPU PLL - 1.88V
VCORE - 1.35V
QPI/Uncore - 1.35V

Failed.


Ninth Try
BCLK - 203MHz
Memory - 8x BCLK
QPI - 18x BCLK
Uncore - 2x BCLK
CPU PLL - 1.88V
VCORE - 1.4V
QPI/Uncore - 1.4V

Failed.


Tenth Try
BCLK - 193MHz
Memory - 8x BCLK
QPI - 18x BCLK
Uncore - 2x BCLK
CPU PLL - 1.88V
VCORE - 1.35V
QPI/Uncore - 1.35V

Failed.



Eleventg Try
BCLK - 193MHz
Memory - 8x BCLK
QPI - 18x BCLK
Uncore - 2x BCLK
CPU PLL - 1.88V
VCORE - 1.4V
QPI/Uncore - 1.4V

Failed.





So my final config is


Fourth Try
BCLK - 183MHz
Memory - 8x BCLK
QPI - 18x BCLK
Uncore - 2x BCLK
CPU PLL - 1.88V
VCORE - 1.3125V
QPI/Uncore - 1.3125V
Stable Overclock of 3.843GHz


From the above I have deduced that I have the worst Core i7 920 chip ever manufactured.
The Chip is hot the worst overclocker.
And I also have an ambient temperature of 37C
So Now it idles at 49C and Max outs on 79C.


So please comment.

Also any suggestions or criticism welcome and desired.
Click to expand...
 
Last edited:
mihir said:
Now I have the True in a push pull and some better airflow.
The Push fan will pull air from outside the case where my air cooler is situated.
and the pull will pull the hot air from the HS and push it towards the back of the case where another exhaust is place which pushes the hot air outside.
So I will be trying for 4Ghz.

I want to know what Memory timings should I keep and what memory clock. :D
thanks


FIRST POST
Click to expand...
That's some serious cooling you got there. :) What kind of temps are you getting with that?
 
apj101 said:
then why are you overclocking?
Click to expand...

I have been video encoding lately so the greater speed helps plus I am finding overclocking really fun.
Its pretty interesting.

Also for benchmarking.



spynoodle said:
That's some serious cooling you got there. :) What kind of temps are you getting with that?
Click to expand...

I am gettin 30C idle.
And I almost took your advice about the refirgerator. :D
 
You must log in or register to reply here.
Back
Top