• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Another C0 -> D0

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

joedymueller

Member
Joined
Apr 6, 2009
Location
Northern Virginia
I just swapped my C0 with a D0 the other day and i have started over clocking it. I got it up to 4Ghz no problems with prime, my problem is with temp. it peaks out at 96 C , i know thats to hot. I am on water, so I dont think it should be that high right?


Do I need to re-seat the water block?
 

doz

Member
Joined
Nov 30, 2002
Location
Las Vegas, NV
No, thats not what it would be. Fans pulling instead of pushing will net no more than about 3c.

You might want to try reseating your waterblock. Also, what are your voltages? D0's seem to run hotter than C0's at teh same voltage (due to lower VIDs possibly?).
 

baditude_df

Northern Senior
Joined
Apr 24, 2007
No, thats not what it would be. Fans pulling instead of pushing will net no more than about 3c.

You might want to try reseating your waterblock. Also, what are your voltages? D0's seem to run hotter than C0's at teh same voltage (due to lower VIDs possibly?).

Absolutely. It would take an awfully long time for your entire loop to heat up to the point where your processor would hit 96deg. Pull fans on a rad will actually do a pretty decent job of sucking air across the fins. Not as efficient as a push fan, but not nearly inefficient enough to cause the temps you're seeing.

As for D0's running hotter than C0's, I haven't seen that parallel personally. If anything they run almost exactly the same for any given clock speed, only the D0's took less voltage to get there. Of course I may be the exception.

Reseat that damn thing and make sure you are not using too much or too little TIM. Just a dot in the middle of the IHS should do the trick. Like So.


And yes, that is way too high. You're roughly 20deg high of the norm, given a 22deg ambient.

Answer us a few questions:
What program to monitor temps?
You should be using THIS.
What is your core voltage? Uncore voltage?
Just realized no loop info in your sig. What does your loop consist of?
 
Last edited:
OP
joedymueller

joedymueller

Member
Joined
Apr 6, 2009
Location
Northern Virginia
the fans are on the inside of the case, pulling air in from the back (into the case). the front fans are on pull (into the case), only leaving the top 140mm fan pushing air out of the case. can't see that being a good thing.



im going to switch the fans on the back around so there is push (out of the case)


Answer us a few questions:
What program to monitor temps? realtemp
You should be using THIS.
What is your core voltage? Uncore voltage? lemme check
Just realized no loop info in your sig. What does your loop consist of? lemme remember.

ill update in the morning.
 
OP
joedymueller

joedymueller

Member
Joined
Apr 6, 2009
Location
Northern Virginia
well i just flipped the back fans that are next to the rad and temps dropped about 3* so far. im going to let prime run some more to see if it hits 100...

below are all my settings.


EVGA VDroop Control – Without VDroop
CPU VCore – 1.47500V
CPU VTT – +300mV
CPU PLL – Auto
DIMM Voltage – 1.65V
DIMM DQ Vref – +0mV
QPI PLL VCore – 1.400V
IOH VCore – 1.400V
IOH/ICH I/O Voltage 1.650V
ICH VCore – 1.250V
PWM Frequency – 1067 KHz
QPI Frequency Selection to “4.800 GT/s.
 

Bobnova

Senior Member
Joined
May 10, 2009
That's rather of a lot of vcore for an i7, seems like most d0's are in the 1.3-1.36 range at 4ghz, i've seen some in the high 1.2s.

I'd work on lowering the vcore.
 

xtkxhom3r

Member
Joined
May 29, 2008
Location
San Antonio,TX
way to much vcore man thats why its burning up...

lower the vcore to about 1.3 and work from there and also take your qpi and ioh down to 1.2-1.35 you shouldnt need 1.4 for 4ghz...

my cpu does 4.0 with 1.26vcore
1.25 qpi
1.24 ioh
 

Bobnova

Senior Member
Joined
May 10, 2009
Essentially, set all the voltages to stock, and turn up the bclk till you have issues, then add a notch or two of voltage and up the bclk again.

Alternatively, you can pick a bclk that you want (like 20x200, for instance), and lower the voltages a step or two at a time till you have issues, then up them a notch till you don't.
 

Norcalsteve

Member
Joined
Sep 19, 2009
Location
Baker, FL
Got this quote from Brutal-Forces post, good input over there: http://www.overclockers.com/forums/showthread.php?p=6360964#post6360964

It may help you A LOT with getting the most out of the lowest voltage possible!

Start by lowering your CPU multi to x16...that way we can keep the CPU out of the picture for a few minutes. To ensure consistency, disable all power saving and throttling features, and set your vcore to 1.25V and enable load line calibration.

Next, back your memory way down to the lowest multi, leave the timings at CL9 how they are now, that should be fine. Lock your vDIMM at 1.65V and your VTT to 1.35V.

Also, change all other BIOS voltages/timings back to "auto".

Now, we are ready to find your basic Bclock limit. Start pushing it up from 180MHz in 5MHz increments with a 5 min stability test each time. If you make it to 210MHz, then drop back to 200MHz and see if the stability test will pass for 30 min. If you cannot make it to 210MHz, try to give it another bump in VTT, to 1.4V or so, and try again.

If you get your Bclock stable at 200MHz with these loose mem timings, the next step is to tighten them up. So raise your mem multi back up (one step at a time) to where you want it...checking for basic (5 min) stability at each step. Although for now, do not exceed the manufactures specs for the memory speed/timings. If you find an instability, raise the VTT a little.

After getting your Bclock and memory stable, we can go back to working on your CPU. VTT & vDIMM are the only voltages we should have played with much so far, because we now have a stable Bclock & mem settings, we can freely OC the CPU and our only variables will be CPU voltage, and heat.

OCing is all about controlling your environment, isolating one component at a time, and pushing each piece to it's limit.
 

Shadowmonk

Member
Joined
Mar 18, 2008
Try these settings

EVGA VDroop Control – Turn LLC on
CPU VCore –1.34
CPU VTT – +700mv
CPU PLL – 1.9
DIMM Voltage – 1.65V
DIMM DQ Vref – +0mV
QPI PLL VCore – 1.375
IOH VCore – 1.3
IOH/ICH I/O Voltage 1.650V
ICH VCore – 1.250V
PWM Frequency – 1067 KHz
QPI Frequency: Auto
 
OP
joedymueller

joedymueller

Member
Joined
Apr 6, 2009
Location
Northern Virginia
I dropped the Vcore to 1.3 and on a short prime run it is stable and doesn't go above 85*

going to play with it some more see if i can get it lower. and will re tim it.
 

baditude_df

Northern Senior
Joined
Apr 24, 2007
1.3v should be enough to power a 4GHz D0 through pretty much anything. Including [email protected], Rosetta or Prime 95.
I've always given my i7's .05v over my minimum stable voltage for insurance that I would not be folding or crunching with CPU that wasn't 100% stable. @4.0GHz, both my 920 and 950 D0 stepping have taken ~1.25v(CPUZ) to remain stable through stress testing, so I set the Vcore to result in 1.30v CPUZ under load. As you can see in the screencap, I am quite a bit lower than that now. I've been experimenting with some new settings which have brought that down a bit to 1.25v CPUZ under load.
Currently it's a nice sunny day in Calgary and my ambient temperature is 26degC. My load temperatures on Large FFT's are averaging ~62deg on water.

With a quality air or water cooler and a good mount, you should be expecting somewhere between 65-75degC load temps at room temperature.
I am willing to bet that your biggest hurtle has been and is going to be keeping that chip cool enough from the get go, to realize it's full potential and to find a nice low vcore to run 24/7. Until you get a good mount, with a proper TIM application and sufficient tension you will likely struggle a bit.
 
Last edited: