Heres my current settings @ 4.533Ghz:
 
Mother Board (EVGA X58 Classified 4 Way 762)
Drivers (Latest)
Bios ( 77 )
CPU ( I7-980X ) (dont know where to check) 
CPU Cooler ( Great water cooling )
Memory ( 12MB 6x2GB Mushkin Redline 1600 )
PSU ( 2x Ultra X4 )
GPU ( EVGA GTX580 Quad SLI )
Drivers ( Latest )
Operating System ( Windows 7 Pro 64 )


Frequency Control
CPU Clock Ratio ( 34X )
CPU Host Frequency (Mhz) ( 133 )
MCH Strap ( Auto )
CPU Uncore Frequency (Mhz) ( 2400Mhz {18X} )
CPU Clock Skew ( 0 ps )
IOH Clock Skew ( 0 ps )
Spread Spectrum ( Disabled )
PCIE Frequency (Mhz) ( 100 )
Turbo Performance - ( Disabled ) 


Memory Feature        
Memory Speed ( Standard )
Memory Control Setting ( Enabled )
Memory Frequency ( 1600Mhz / 2:12 )
Channel Interleave Setting ( 6 Way )
Rank Interleave Setting ( 4 Way )
Memory Low Gap ( 2.5G )
tCL Setting ( 6 )
tRCD Setting ( 8 )
tRP Setting ( 6 )
tRAS Setting (Auto ) "29"
tRFC Setting ( Auto ) "88"
Command Rate ( 1t )


Voltage Control
EVGA VDroop Control ( Without VDroop )
CPU VCore ( 1.44375 )
CPU VTT Voltage ( +200mv )
CPU PLL VCore ( 1.845v)
IOH PLL VCore ( 1.845v ) 
QPI PLL VCore ( 1.400 )
DIMM Voltage ( 1.65 )
IOH VCore ( 1.175)
IOH/ICH I/O Voltage ( 1.650v )
ICH VCore ( 1.175v )
NF200 Voltage (Auto) "1.200" 
VTT PWM Frequency ( 490 KHZ )
CPU PWM Frequency ( 1210 KHZ )
CPU Impedance (Auto)  

CPU Feature
Intel SpeedStep ( Disabled )
Turbo Mode Function ( greyed out )
CxE Function ( Disabled )
Execute Disable Bit ( Disabled )
Virtualization Technology ( Disabled )
Intel HT Technology ( Enabled )
Active Processor Cores ( All )
QPI Control Settings ( Enabled )
QPI Link Fast Mode ( Enabled )
QPI Frequency Selection ( 4.800 GT/s )
OC Recorvery ( Disabled )

PnP/PCI     PCI Express
Maximum Payload Size ( 256 )
 
Please give me some suggestions.. Would obviously love something over 4600.. 4700+ would be sweet.
 
I've tried going faster but Prime 95 will just lock right away if i go any higher.. I've tried upping the cpu vcore up to 4.7 and that didnt help.. upped the VTT +50 and that did nothing. I've also tried removing a set of ram and just runing it with 6GBs thinking that was forsure what was holding me back and nope! Same thing.  Plus tried lowering the timings to 8,8,8 2T and that didnt make a difference either.
 
I've run out of ideas but i still believe it should get atleast 4.6Ghz.
 
 

The Main thing I notice you're doing wrong is  your Dram multiplier(2:12) and uncore frequency (x18). You said you messed with the memory timings, but what about the speed? The memory's speed directly reflects your Uncore frequency. So what you should do is lower the speed of the memory and uncore to its lowest possible setting, while keeping a high cpu clock. Best way to overclock a cpu is to take the memory out of the equation and just work with the cpu.
 
Some of your voltages look like they may be too high.
These 3 in general...
CPU PLL VCore ( 1.845v)
IOH PLL VCore ( 1.845v ) 
QPI PLL VCore ( 1.400 )
You could actually gain stability by lowering these, but i dont know your chip, maybe you've already tested lower and you know you need these settings.
 
 
 

Wait a sec.. must not of tried 1.47v cause when i went back into the bios it showed 1.46.. so i changed it to 1.475v and it lasted a whole 20 secs on Prime 95 before it locked up and the same BSOD.
 

Oh wait a sec this is the guy with the 4 580s wc'ed that could give me a run for my money....no more info for you.....j/k but no really 4.8 is all u get I can't tell all my secrets. 
Nice rig btw
 
Also be aware that as you OC whether or not your passing your max ram frequency your going to have to loosen the timing.

Alucard666
Any voltage over 1.35 is a no go for me. I'd steer clear of anything needing so much vcore.

I have to agree that except for benching or short-term use I would not push clocks requiring such high Vcore.  I'm actually running at 4.015GHz on the i7 970 for 24/7 at nice low voltage of around 1.23V at load.

I thought when the 980x came out you were not suppose to go over 1.4v on Vcore because of the internal memory controller.  Although I see a lot of people using more then 1.4v on Vcore and seem to be ok. I run 4.4ghz on 1.399 according to EVGA E-leet. 

OV3RCLK4

EleSSaRBLS

I have pushed this cpu to 5.4 ghz on 1.66vcore on straight water with no problems. Now that I look at my notes I have achieved 4.8 on 1.387 vcore but it wasn't stable for a long duration. I'm not saying he should or shouldn't but for these chips 1.48 isn't that high and will only render max temps in the high 50's low 60s. With a good wc setup those high clocks are available. Your not going to reach 4.5 or above on an h50 or h70 at no ammount of vcore w/out max temps going through the roof. I guess it comes down to what your comfortable with. I get 5+ out of mine cause I'm willing to push it. My max temps even at 5.2ghz with my wc setup only his mid 60's which is well within safe operating temps. A cpu is only dependent on the temp you can keep it at. If I could keep it at 32c I could lay 2.0v vcore to it and it wouldn't matter.

That part I put in BOLD is completely false. If Lephron was around he would make a nice long reply on what Vcore does. There is a math equation for it.  Even if you do run 2.0Vcore or 1.99Vcore whatever, and the CPU is on Liquid Nitrogen -50c during load, there is still WAYYYYY TOO MUCH VOLTAGE running through the cpu, this will severely degrade the life time of the processor. This doesnt mean your chip will instantly fry, but a chip that could have lasted 15 years, will now only last 2 years or even less.

To the OP, Please try what me and Dejanh said. Lower the CPU PLL and IOH PLL, and also lower your memory speed to 2:8 and uncore to x15. THEN, try upping the Vcore and VTT to get more Mhz for your CPU overclock. You are already pushing it plenty and cutting the life of your processor in half. Thing is, you will upgrade before it dies anyway because these chips can last 10+ years when not overclocked.

+1
 
I did not feel like commenting earlier
 
It bugs the heck out of me when I see people thinking they can push unlimited amounts of voltage through their chips and achieve unlimited scaling.  There is no such thing and no amount of cooling will help you.  Even running at 1.5V for 24/7 will severely degrade the life of the chip, but for most of us this will not be an issue before we replace it anyway.

Oh and btw.. my temps were reporting low 60s to high 60s across all cores at 1.475v on load.. I've got a ton of rads and a ton of pumps but not the fastest loudest fans like i used to.. just GT 1850RPM fans in push/pull. Abundant temps are around 77F in the room.
 
 

sgtharry

For whatever it's worth, I have been running mine with 1.575 v-core for almost a year.Thats what mine takes to do 4.5 stable. No ill affects so far, and my temps never top 70 while running LinX. Some 980's need the v-core.

 
What VTT are you runing as well?? Also are you using V-Droop?

My previous CPU i7 930 ran 24/7 at 4.4Ghz.. I upgraded it to the i7 980x expecting a number north of 4.6Ghz to justify the upgrade.
 
I'm not looking for 5Ghz.. if i was i would not of bought the case i have. I just wanted a sweet lastest and greatest gaming rig.. Benching was the bonus.. and why not after spending all that cash on it. 
 
I just want it runing where it should be so i can settle and start playing games as intended. Though i cant rest if theres a problem and wont till this overclock is resolved.
 

dejanh

EleSSaRBLS

Falcon_CMH

dejanh

Falcon_CMH

Qpi pll 1.45
CPU pll 1.9
And plenty of vcore

You do not bump CPU PLL on these chips, you lower it.  1.5V and below increases stability with Westmere arch. chips.

@AHowes - try dropping CPU and IOH PLL values for more stability.  Even with old Bloomfield chips this held true to a great extent.  I never ran my CPU and IOH PLL on any of my 4C/8T i7 9xx series chips above 1.5V for best voltage and stability.

CPU PLL at 1.9 is correct I have my 980x running at 5Ghz.  The higher the CPU PLL the lower you can run the vcore but 1.47 is way to low.  You need 1.65v+ to get in the 5Ghz range.

And how is it I run linx stable @ 5.2ghz 20 passes 25k problem with a mere 1.58 vcore and as far as life who cares really. I'm not saying it should be ran at 4.8 for its entire life as I don't do that to mine, just for benching, linx and other benchmark #'s.  I will say I may have exaggerated 2.0 vcore it was a point that the ammount of voltage is dependent on how cool you can keep it. I have seen many runs at 1.89+ on LN2. Can something burn out sure it can, but it can anytime you go above intels spec of 1.38. I was simply giving the guy what works for mine, but as we all know all chips are not the same. As far as the pll and cpu vcore it has to be dropped, but I prolly don't know much about overclocking.

I want to see this, honest, I do.  5.2GHz with 6C/12T using LinX with 10.3.0.003 libraries (download here http://www.mediafire.com/?ubhp560c83nwkfv) and all DRAM running 50-pass (minimum problem size should be no less than 25K)...you can do that you got one very, very nice chip.  RealTempGT 3.6 monitoring temps of course, idle and load shown No tossing your computer out the window

@OP - I need 1.47V at load roughly to run 4.62GHz on the i7 970.  My IOH and CPU PLL are below 1.3V for stable runs at that.  I did not spend much time tweaking so I could probably shave down the volts just a bit, but maybe this will give you some idea what to try.  Also, you are running with all 6 slots of DRAM.  That is putting a lot of stress on the chip and in general will limit your overclocking ability.  Doing 4.6GHz with 3 slots and 6 slots is not the same thing at all.

 
Here ya go eat your heart out....
 
 
Re:*Club i7 Stability Continuation* Monday, January 10, 2011 2:07 PM ([link=javascript:void(showMsgNum(784570))]permalink[/link])
Heres with the updated linx binaries. I accidentally had it on 25 passes so i went ahead and stopped it after 21.
 
EleSSaRBLS 
Core i7 (980x)   Batch #:  3032A340
CPU clocked at: 5.2 Ghz 
Voltage:  1.59 Peak
Idle temps: -5 -5 -2 -13 5 -3 
Load temps: 56 54 51 42 59 54
 

<message edited by EleSSaRBLS on Monday, January 10, 2011 2:14 PM>

 
 
70,133 3dmarks: http://3dmark.com/3dmv/2860484
34,839 3dmark06: http://3dmark.com/3dm06/15266493    
E762 4X SLI Classified
I7 980x 4.4ghz everyday
6gb Corsair Dominator GT 2000mhz  
Corsair AX1200 Gold PSU  
3x 120Gb OCZ VERTEX 2 RAID 0 
4x EVGA GTX 480SC 940/1880/2180 

I think 20 passes as I stated should suffice.

dejanh

I took my discussion with you offline.  Let's just say I'm not satisfied at all.

 
Flocko puts popcorn in micro and waits !!!

Voltage isn't everything, temperature is however, following along the basic concept relating to how the resistence of a circuit varies with temperature.  I'd suggest finding ways of getting your full load temperatures lower before thinking more voltage will fix it, which just raises temperatures even more.  World record runs are done not just due to the high voltages applied, rather the extreme temperatures achieved, many in cryogenic territory.
 
For a formula sense, it is a square root function where you will need a greater temperature delta from the maximum for stock to account for the increase in wattage ratio.  The very idea that overclocked tmepratures can exceed stock max temperatures or even reach the same value is just negligent.  Just because nothing bad happens sooner doesn't make it right, just folks are lazy, they want the results of a high overclock (fps or scores) than processor integrity.
 
The whole reason we up the votlage is that more electrons in the circuit escape when ramping up the frequency.  The science behind this is called electromigration, which begins as soon as a stock processor is powered up.  Electromigration is what determines how long a processor can last, not necessarily whether it is overclocked or not.  Over time, a processor will loose electrons and need a higher voltage just to run stock towards the end of its life.  This is why the Vcore is set high enough to give us the impression of good overclocking overhead for shorter periods than the warranty, there is actually more voltage needed for stock than when you first fire it up.  Whatever is required for Intel's 3yr non-stop 24/7 warranty.
 
Trouble is how we define "stability".  People think if after running some random extreme stress test for hours or days, they call it stable -- no it isn't.  It could give in a week or a month and it depends on the frequency set, or the RAM or mainboard circuitry could give -- we test total system stability when we run these tests, not whether one part can take the punishment or not.
 
For instance with CPU overclocking, is all the voltage only going straight to CPU cores and/or caches?  Not in Nehalem on up, which has controllers and whatnot that have to deal with power and signal coming in faster than it was certified to take.  It is a major gamble to want more frequency and I think folks with good batches are just lucky and shouldn't expect more than they get.  Overclocking it isn't (nor should it be, IMO) easy or simple just because mainboard and heatsink makers give us the options to do so.

Thanks for the info.. but i'm aware of most of that as i've been overclocking computers since the Pentium 2 chip came out. heh.
 
Just was wondering if there was something i was doing wrong.. if i've reached the most i can get on water then i'm done.  As i said, this is my gaming rig and just wanted it runing at its max in a beautifull case.. I'd of built a bench and saved a ton of dough if i wanted to do phase change or LN for extreme benches.
 
In the next hour i'll try removing 3 sticks of ram and give it a go again with the new info once my wife gets home and my almost 2 year old son and almost 6 year old daughter is asleep.
 
If in the end it responds well to 5 volts then i'll know its voltage thats holding me back. As i said 4.6Ghz and i'm happy.. didnt think i'd have to raise the voltage from 1.43 to 1.5 to go another 100Mhz.
 

AHowes
didnt think i'd have to raise the voltage from 1.43 to 1.5 to go another 100Mhz.

You may just...scaling is not linear
 

EleSSaRBLS
and basically supplementing what I said that the lower the temps you can maintain the higher you are able to overclock and the more voltage you are able to apply due to the fact that there is less resistance and the electrons are able to move more freely in the silicon channels to and from their desitnation. Voltage although it does create heat does not create no where near the heat that resistance does.  

Not voltage, current.  And temperature results in increased resistance