Take your time . I could hit, I think it was 4.3GHz (207) on air. You should be able to do a little better running H20. Have fun...

Heres the BIOS template for my 4.2 for those of you who can look at it and give some suggestions to achieving 4.4-4.5 while I try to work on it as well...Thanks everyone


Frequency Control
CPU Clock Ratio ( 21x )                                                                                         
CPU Host Frequency (Mhz) ( 200 )                                                                      
MCH Strap ( 1600 )
CPU Uncore Frequency (Mhz) ( x16 )                                                
CPU Clock Skew ( 0 ps )
Spread Spectrum ( Disabled )
PCIE Frequency (Mhz) ( 100 )


Memory Feature        
Memory Speed ( standard )
Memory Control Setting ( Enabled )
Memory Frequency ( 2:8 1600 )                                                                
Channel Interleave Setting ( 6 Way )
Rank Interleave Setting ( 4 Way )
Memory Low Gap ( Auto )
tCL Setting ( 26)
tRCD Setting ( 8 )
tRP Setting ( 8 )
tRAS Setting ( 24 )
tRFC Setting ( 74 )
Command Rate ( 1t )


Voltage Control
EVGA VDroop Control ( W/O VDROOP )
CPU VCore ( 1.3175 )                                                                                                
CPU VTT Voltage ( +125 )
CPU PLL VCore ( Auto)
**IOH PLL VCore (Auto ) 
DIMM Voltage ( 1.65 )
DIMM DQ Vref ( +0 )
QPI PLL VCore ( Auto )
IOH VCore ( Auto )
IOH/ICH I/O Voltage (  Auto)
**BR04 ( NF200) Voltage ( )
**VTT PWM Frequency ( 250 )
**CPU PWM Frequency ( 800 )
**CPU Impedance ( LESS )
**QPI Signal Compensation ( LESS )
ICH VCore ( Auto )
*PWM Frequency ( )


CPU Feature
Intel SpeedStep ( )
Turbo Mode Function ( ENABLE)
CxE Function (  )
Execute Disable Bit ( DISABLE )
Virtualization Technology ( ENABLE  )
Intel HT Technology ( ENABLE )
Active Processor Cores ( ALL )
QPI Control Settings ( ENABLE )
QPI Link Fast Mode ( ENABLE )
QPI Frequency Selection ( AUTO )
OC Recorvery ( ENABLE )

 
This makes no sense... my time per pass is lower and my GFlops are higher than allwalleyeguy's yet his overall time is faster???  Is his CPU throttling? All in all LinX is all over the place with no real consistency.  I'm not sure if OCCT would be better... it does have LinPack and graphs which would help.  How about Intel Burn or just could old fashion Prime95 which is quite easy on the cpu?
 
 
 

overclocker333 This makes no sense... my time per pass is lower and my GFlops are higher than allwalleyeguy's yet his overall time is faster???  Is his CPU throttling? All in all LinX is all over the place with no real consistency.  I'm not sure if OCCT would be better... it does have LinPack and graphs which would help.  How about Intel Burn or just could old fashion Prime95 which is quite easy on the cpu?

Two differences. 1.) LinX 6.3 V. 6.4, and 2.) and this is could be a biggy, memory speed. One way to stay cooler is to run the memory slower.
[edit] Throttling would not change one way or the other no matter what test vehicle is used. If a processor throttles as a function of temp, then no matter how i gets hot it would still throttle. Also, and to your point, simply because a device displays less than 100c in Realtemp, it may have had to throttle to get back down to an allowable temperature.

Frequency Control
CPU Clock Ratio ( 21x )                                                                                         
CPU Host Frequency (Mhz) ( 200 ) 210                                                                      
MCH Strap ( 1600 ) I set to 1867
CPU Uncore Frequency (Mhz) ( x16 )                                                
CPU Clock Skew ( 0 ps )
Spread Spectrum ( Disabled )
PCIE Frequency (Mhz) ( 100 ) 101-102


Memory Feature        
Memory Speed ( standard )               I loosen memory timings til I get O/C I want then tighten them up til it dies.
Memory Control Setting ( Enabled )
Memory Frequency ( 2:8 1600 )                                                                
Channel Interleave Setting ( 6 Way )
Rank Interleave Setting ( 4 Way )
Memory Low Gap ( Auto )
tCL Setting ( 26)
tRCD Setting ( 8 )
tRP Setting ( 8 )
tRAS Setting ( 24 )
tRFC Setting ( 74 )
Command Rate ( 1t )


Voltage Control
EVGA VDroop Control ( W/O VDROOP )
CPU VCore ( 1.3175 )  +++ as needed                                                                                              
CPU VTT Voltage ( +125 ) 175
CPU PLL VCore ( Auto) 1.425
**IOH PLL VCore (Auto ) 1.425
DIMM Voltage ( 1.65 )
DIMM DQ Vref ( +0 )
QPI PLL VCore ( Auto ) 1.2
IOH VCore ( Auto ) 1.15
IOH/ICH I/O Voltage (  Auto)
**BR04 ( NF200) Voltage ( )
**VTT PWM Frequency ( 250 ) up to next setting
**CPU PWM Frequency ( 800 ) ditto
**CPU Impedance ( LESS )
**QPI Signal Compensation ( LESS )
ICH VCore ( Auto )
*PWM Frequency ( )


CPU Feature
Intel SpeedStep ( )
Turbo Mode Function ( ENABLE)
CxE Function (  )
Execute Disable Bit ( DISABLE )
Virtualization Technology ( ENABLE  )
Intel HT Technology ( ENABLE )
Active Processor Cores ( ALL )
QPI Control Settings ( ENABLE )
QPI Link Fast Mode ( ENABLE )
QPI Frequency Selection ( AUTO )
OC Recorvery ( ENABLE )                            My .02

After looking through the latest results posted, I'm puzzled by the results posted by AWALLEYEGUY. By looking at the overall time and adding in the rest period between each run, his results should read out at close to 1hr 6min rather than the 1hr 3min it shows. The math doesn't work out. By default LinX has the same rest period between each run before placing into full load status again. This is with all systems. AWALLEYEGUYs numbers indicates that the rest time with his run was shortened by almost 50% thus shaving 3min from his time. I'm getting a feeling this is another bug with LinX and it's timer. I've run LinX where the clock has paused or stopped for a few seconds upto a minute in some cases then resumes without compensating the time for the period paused or stopped. This inconsistancy can also be seen in other postings.
  
Also noticed that there seems to be a uneven loading while running LinX from one core to the next. If load was equal all the way across then temps for each core should also be equal for 100% load. In all my runs as well as what I see with other post, this is not the case and a difference of almost 15c can be seen from core 0 to core 3. My own temps show a difference of 10c between the 2 cores. This means load per core and threads is inconsistant and unequal. This also I think is a bug with LinX like the bug where if it finds a little instability, rather than stopping, it will throttle itself back and continue to run making for inaccurate results.
 
Just thought I would point this out.

Frequency Control 
CPU Clock Ratio ( 21x )                                                                                          
CPU Host Frequency (Mhz) ( 200 ) 210                                                                       
MCH Strap ( 1600 ) I set to 1867 
CPU Uncore Frequency (Mhz) ( x16 )                                                 
CPU Clock Skew ( 0 ps ) 
Spread Spectrum ( Disabled ) 
PCIE Frequency (Mhz) ( 100 ) 101-102 


Memory Feature         
Memory Speed ( standard )              
Memory Control Setting ( Enabled ) 
Memory Frequency ( 2:8 1600 )                                                                 
Channel Interleave Setting ( 6 Way ) 
Rank Interleave Setting ( 4 Way ) 
Memory Low Gap ( Auto ) 
tCL Setting ( 2627) 
tRCD Setting ( 8 ) 
tRP Setting ( 8 ) 
tRAS Setting ( 24 ) 
tRFC Setting ( 74 ) 
Command Rate ( 1t ) 


Voltage Control 
EVGA VDroop Control ( W/O VDROOP ) 
CPU VCore ( 1.40 )                                                                                               
CPU VTT Voltage ( +125 ) 175 
CPU PLL VCore ( Auto) 1.425 
**IOH PLL VCore (Auto ) 1.425 
DIMM Voltage ( 1.65 ) 
DIMM DQ Vref ( +0 ) 
QPI PLL VCore ( Auto ) 1.2 
IOH VCore ( Auto ) 1.15 
IOH/ICH I/O Voltage (  Auto) 
**BR04 ( NF200) Voltage ( ) 
**VTT PWM Frequency ( 250 ) up to next setting 
**CPU PWM Frequency ( 800 ) ditto 
**CPU Impedance ( LESS ) 
**QPI Signal Compensation ( LESS ) 
ICH VCore ( Auto ) 
*PWM Frequency ( ) 


CPU Feature 
Intel SpeedStep ( ) 
Turbo Mode Function ( ENABLE) 
CxE Function (  ) 
Execute Disable Bit ( DISABLE ) 
Virtualization Technology ( ENABLE  ) 
Intel HT Technology ( ENABLE ) 
Active Processor Cores ( ALL ) 
QPI Control Settings ( ENABLE ) 
QPI Link Fast Mode ( ENABLE ) 
QPI Frequency Selection ( AUTO ) 
OC Recorvery ( ENABLE )                            






This is what I did...SOOOO...I ran LinX and the first run stopped with errors...didnt freeze and didnt BSOD this time!!! So we are getting close!!!


What now?

awalleyeguy

expedision

@ awalleyeguy I have always ran MCH Strap in auto, when did you find a need to change this for stability . I never messed with it, but never oc as far as you have either.

I was reading somewhere, and thought I would give it a try did not seem to hurt anything. Will not boot at 2133 though.

---

@ trs32505 I have noticed with linx time and realtemp time do not run the same.

 
This true. This why I feel there is a bug with LinX and causes to many inconsistancies with final results.
 
As for Strapping. This affects what timings the chipset will use with memory when memory timings are set to AUTO. Will have a big affect if not set properly.

expedision

So am i correct in saying, it's a setting that should not be messed with. I have ran in auto up to 4.6ghz with no ill effects.

This is correct to a point. It comes in handy when pushing your memory to high speeds. It allows you to set looser timings with just 1 keystroke rather than manually setting each timing induvidually.

I would, you can always tweak later. memory I know is a big factor  so because of that I always try to eliminate from the equation of O/C failing.

Robbyn16

sorry to interrupt your convo guys but did i make the cut for the club?!? haha i want the 4+ sig!!!

If you legitamatly got to the 4.0 mark then go ahead and add the sig :-) I dont have time due to finals right now to update the list, so go ahead...lol



And back on topic of this OC im working on...I tried to bump up those voltages suggested and now it actually crashes rather than just error out on LinX which is strange to me...the vcore is in the right place, just have to get the others to cooperate now...any other ideas? I tried to loosen the timings up to 9-9-9-24-74 and that crashed as well...

Keep the help coming!!! I wanna be up there with yall!!!


SimC

Hi guise. I just built this new rig 3 days ago. What I did here is bump everything to near safe max (voltage-wise) and see if it can do 4.0 ghz. I got a lot more tweaking to do, but I'm happy I got 4.0 :D

Core i7: 920   Batch #:  3939B215
CPU clocked at:  4001.0
Voltage: 1.399 (idle) 1.376 (load) Vdroop ON
Idle temps:  41, 38, 40, 37
Load temps:  86, 83, 81, 80 (still curing in AS5)
HT/Turbo = ON
Megahalems + 2750 RPM Panaflo

chriscbenjamin

when am I allowed to put the 4+GHZ stronger than steel, faster than lightning pic in my sig? Should I wait till I am officially added to the list?

 

As I posted earlier...If you legitamatly got to the 4.0 mark then go ahead and add the sig :-) I dont have time due to finals right now to update the list, so go ahead...lol 




I just looked at your post of your OC and your good to go...add the badge 

mudvari

hey sim good to see you back...!! hope you reach 5GHz OC...and for that may the lords and devils come together and shower you with their choicest blessings!!

as for me I was able to reach 4.3GHz but without HT...I had a problem with my RAM so had to RA it...and had to revert back my OC to 4.2GHz with HT on...:) !!!

lol Thanks for the blessing of the lords and devils but I'm not going for 5ghz lol...theres no way in hell my chip is gonna allow that. I just want to stop at 4.4-4.5...

What RAM are you going to be using? I was thinking about changing mine out...


SimC

Here's a modest 4ghz effort just to get a badge! lol
" border="0" />