I cannot get it stable at 425x9 any1 can guide me with voltages please? my mobo is an asus maximus extreme using corsair H-50 water cooler!
Printable View
I cannot get it stable at 425x9 any1 can guide me with voltages please? my mobo is an asus maximus extreme using corsair H-50 water cooler!
you need to change your username to Q9650,i think your new cpu is upset:D
seriously though,the x38 isnt a great board for these quads,but still you should post
your bios settings for people to help.also is the maximus extreme an x38 board?
My current 3.6ghz OC 400x9 bios settings on the maximus extreme
Ai Overclock Manual
CPU Ratio Setting 9
FSB Frequency 400
FSB Strap to North Bridge Auto
PCIE Frequency 100
DRAM Timing Control Manual
DRAM Frequency DDR3-1600
DRAM Command Rate 1T
CAS# Latency 7
RAS# to CAS# Delay 6
RAS# Precharge 6
RAS# Active Time 20
RAS# to RAS# Delay Auto
Row Refresh Cycle Time Auto
Write Recovery Time Auto
Read to Precharge Time Auto
Read To Write Delay(S/D) Auto
Write to Read Delay(S) Auto
Write to Read Delay(D) Auto
Read To Read Delay(S) Auto
Read To Read Delay(D) Auto
Write To Write Delay(S) Auto
Write To Write Delay(D) Auto
DRAM Static Read Control Disabled
DRAM Dynamic Write Control Disabled
Ai Clock Twister Strong
Ai Clock Skew for Channel A Auto
Ai Clock Skew for Channel B Auto
Ai Transaction Booster Enabled
Boost Level 3
CPU Voltage 1.15 vdrop 1.12
CPU PLL Voltage 1.50
North Bridge Voltage 1.35
DRAM Voltage 1.80
FSB Termination Voltage 1.2
South Bridge Voltage 1.075
Loadline Calibration Disabled
CPU GTL Reference 0.63
North Bridge GTL Reference 0.67
DDR3 Channel A REF Voltage Auto
DDR3 Channel B REF Voltage Auto
DDR3 Controller REF Voltage Auto
SB 1.5 Voltage Auto
CPU Spread Spectrum Disabled
PCIE Spread Spectrum Disabled
ok im not 1005 familiar with the x38 but have used the x48 and i do know the NB
voltage required for my quad at higher fsb was a lot.i would start with loosing up
your ram timing and focus on the cpu first. try more FSB term voltage-1.26v and cpu pll-1.57v
give it a little more vcore and test to see if its stable and then start dropping vcore.also try 1.42v NB
I have that same combo in my setup now under water. That Q9650 will go higher under that board if you give the NB the juice it needs. Mine operates in the 1.57V area to keep it stable crunching 24/7 at 4.05 GHz. Then again it's also watercooled and not using the stock cooling. My CPU Vcore is at 1.28 if I remember correctly (I'm at work right now).
yea thats what i was thinking too.i couldnt remember the NB voltage needed for my quad
on my rex but i did remember it was a lot more than the P45 chipset.
New overclock!!!!!:shocked:
3.8ghz 425x9:rocker::worship:
Vcore in bios is 1.300v
x38 voltage is 1.61v
http://img812.imageshack.us/img812/9318/46280462.png
that cpu can go much higher, mark my words :P
I am worried about load temp it gets to 65degrees ! summer heat is holding me back! I can reach this OC thanks to my corsair h50 cooler! but with a vcore of 1.256volts at 3825mhz is simply impressive!!!!!!!!!
Just a heads up for all you 775 Quad o/cers out there (and dualies) - all 3 of my P45 boards are up for sale - all in great condition and fully functional:
http://www.xtremesystems.org/forums/...d.php?t=255084
I'm open to offers btw..
Q6600 - looks like you need a better board, you could probably go even further with lower volts on the USB3P, and you'd get to keep you DDR3! ;)
hey fellas,
not been OC'ing for a few months - hungry for more!!!
any pointer??
Q9650 @ 4.95Ghz 1.275v
vFSB - 1.26v
9*455
DDR2 @ 1093 5-5-5-15 2.14v
gtl 1 - 0.675v
gtl 2 - 0.675v
pll - 1.56v
vNB - 1.26v
vSB - 1.1v
LLC - enabled
All spectrum controls - off
asus P5Q deluxe on 2301 bios
it took ages for me to get this OC stable - when i try an boot at 460 i get problems - i'm fully liquid cooled - so i'm not too bothered about high vCores -
i'm thinking - 470@1.3v vFSB : 1.28 and move my ram down to strap below as its only 1066
any tips??
update ... i'm now stable @ 4.212Ghz :) vCore @ 1.325v vFSB @ 1.3v pll @ 1.56v vNB @ 1.28v thats 468*9
boosted my 3d mark 06 cpu score up from 4.095Ghz by about 450 points!!
cheers!!! hey - i see in your sig you've got a q9650 @ 4.8Ghz - what board and what PLLvoltage u got?? do you think my 1.56v PLL is too high?? thats the only voltage i'm concerned about - i'm fully liquid cooled but i dont want to fry the circuits !!
any tips - whats your PPL voltage? and what gtl refs u use?? i'm on x0.675v for both - seems to work better and give higher Gflops output that when i had them on 0.667..
With 45nm ~1.60 vPLL is the soft limit, so keep it under that, preferably under 1.57v
Pretty sure NoGuru used a UD3P for those clocks, its widely understood to be the best board for overclocking quads.
Raw CPU stability (Prime95 small FFT's, LinX) is pretty easy to do on most boards up to a point, its Prim95 large FFT's and long blend runs that really expose instability. You may find it tough to pass especially with 4GB+ ram installed. You would want to run a few hours large FFT's or blend run to be sure you're stable.
thanks cryptik - gonna say - for a 533mhz fsb on nogurus quad - its got to be a gigabyte!
i use 20 loops of linx - then game really heavily for a weekend - i recon 20 loops linx is enough games is where i've found all my instabilites with all my other cpus - when p95 was 8 hours stable games BSOD - so since then i just do
20 loops linx and hardcore game for a few days!!! good enough for me!!
Sure whatever works for you man, stability is subjective anyway. Just passing on what a few of the Q9650 boys in here (and myself) prefer to do but it's all just personal choice of course. Most guys run vPLL between 1.50v -- 1.60v but lower the better, only raise a voltage if you need to!
.785 vREFs - i'm currently using .675 for my OC seems to like it and i've only got a vCORE of 1.325v
The ud3 reads refs differently from the p5q so you cant compare the numbers directly, the 0.785 on the ud3 is about the same as the .655 on the p5q and the 0.800 is about .665-.670 on the p5q.
Guys what is the best batch# of Q9650?
There are a few good ones, and there seems to be some variation between the chips within a batch. Have a look through this thread, there's lots of good examples.