The Official OC (OverClocking) Thread!

That chip's watercooled. 44C on cores means a main CPU temp of no more than 35. At 1.42V, it wouldn't idle at 35, let alone run like that under load.

 

sam,
It's also not going to last too long at that voltage either. Even being water cooled, Electro-Migration has already started! While heat does hasten the process, the primary issue is totally electronic, and mathematically predictable! All water cooling can do is slow down the process. LN2 will slow it even more. Error correction in the CPU itself will handle it for now, but eventually the errors will become greater than the ECC can handle, and that's where it gets real interesting, because you'll have no idea from one day to the next, what the hell is wrong with your computer. One day you'll think it's the memory because it fails Memtest with thousands of errors. Yet checking the same memory in another computer yields no errors at all! With my E6750, I knew it was hurt, shortly after I bumped the voltage to 1.576 to get to 4GHz when I first got it, because it never came close to that speed again. It wouldn't do a lick over 3.55GHz after that and be completely stable, and I kept having to add a little voltage every 4-6 weeks to keep it stable. When I ran out of adjustment for the voltage, it just went nuts. I had lock-ups, some requiring the PSU to be shut down to turn the computer off, system re-boots, and File corruption errors on a CD (impossible to have)! It also started to corrupt files all over my hard drives. If I hadn't had backups of all my software, I would have lost about 1/3 of what I own. I had about 150 or more programs stored in my Software folder, and the ones that were corrupted, were trashed! It took me about 2 weeks to figure out that the problem was the CPU. I finally put my spare, brand new D-940 in it and overclocked it to 3.7GHz and all the problems vanished! I had the replacement E6750 Intel sent me up to 3.8GHz with a much lower voltage than it took my hurt one to get to 3.55GHz. Of course I could have lived without those couple of weeks I spent tearing my hair out trying to figure out what was wrong with the damn thing! LOL!!

I might have been the first here on AD to realize what my problem was, but I think Rob was the first to experience Electro Migration with his E8400. In fact, his experience with the things that were happening to his computer, is what finally convinced me to RMA the CPU. If you remember he changed pretty much everything with that build, until there was nothing left but the CPU, and when he changed that all the problems vanished. The most maddening part of the experience is you would short the CMOS and set it all up, and it would run beautiful for a day, day and a half and then wouldn't boot up the next time you turned it on, until finally it wouldn't boot the computer because of a corrupt config file error. If you tried to boot from a floppy or CD, it would pop up the no NTDLR (or something like that) was missing. Time for Killdisk II! LOL!!

That Q9650 will fail! It's not a question of if, it's simply a question of when! Whether it will be covered by warranty or not is a good question. I know that on the higher end CPUs, they do look more closely, and they can tell if it's been run past maximum volts (1.3625v for the Q6750) and determine if the problem is caused by Electro Migration. I have a Dentist friend of mine with a SkullTrail platform, socket 771, and he didn't realize that the voltage wasn't the same for the QX9775 as the socket 775 QX9770 and fried the CPU. The QX9775 has no voltage range, just one voltage, 1.212v! If you exceed it and it blows, Intel will not replace it! You will have to buy a new one! With LN2, Dr Lee had his clocked to just a hair over 6.0GHz at a shade above 1.9v when it quit. It just stopped, never to run again! I personally saw it run at 5.7GHz on LN2! You just got to be nuts, if you ask me! LOL!!

Best Regards,
Russ

 

I use that voltage with my Core 2 Quad. 1.4 to 1.425V is what I've used a lot. I have had it higher than that as well temporarily.
I'm well aware of the issue with electronmigration. 175mV is the gnerally regarded 'upper limit', it isn't usually a breach of the safety zone. If I was using 1.5V, I'd be worried.
Additionally, the 'no voltage range' QX9775 CPUs in Zardon's system at Driverheaven have been running since summer 08 at 1.4V with no issues.

 

So I have been trying to overclock this processor with no great success:

So far:
400 x 8 = 3200mhz (unstable)
333 x 9 = 2999mhz (stable)
400 x 9 = 3600mhz (unstable)
378 x 9 = 3400mhz (unstable)

with the following settings, can anyone tell me what to do in order to get that 400 x 9 = 3600 stable???

PC Health Status
Adjust CPU Temp: Auto

CPU Feature
- Thermal Management Control: Disabled
- PPM(EIST) Mode: Disabled
- Limit CPUID MaxVal: Disabled
- CIE Function: Disabled
- Virtualization Technology: Disabled
- Core Multi-Processing: Enabled

Exist Setup Shutdown: Mode 2
O.C. Fail retry Counter: 0
O.C. Fail CMOS Reload: Disabled
CPU Clock Ratio: 10x
CPU N/2 Ratio: Disabled
Target CPU Clock: 3330Mhz
CPU Clock: 333
Boot Up Clock: AUTO
CPU Clock Amplitude: 800mv
CPU Clock0 Skew: 0ps
CPU Clock1 Skew: 0ps
DRAM Speed: 333/667
- Target DRAM Speed: DDR2-667
PCIE Clock: 100mhz

CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled

Voltage Settings
CPU VID Control: 1.1750v
CPU VID Special Add: AUT0
DRAM Voltage Control: 1.92v
SB Core/CPU PLL Voltage: 1.51
NB Core Voltage: 1.465
CPU VTT Voltage: 1.100
Vcore Droop Control: Enabled
Clockgen Voltage Control: 3.45v
GTL+ Buffers Strength: Strong
Host Slew Rate: Weak
x MCH RON Offset Value:
x MCH RTT Offset Value:
x MCH Slew Rate Offset Value:
x MCH VREF 1 Value:
x MCH VREF 2 Value:
x MCH VREF 3 Value:
x CPU GTL 0/2 REF Volt: 0.667X
x CPU GTL 1/3 REF Volt: 0.667X
x North Bridge GTL REF Volt: 0.61X

DRAM Timing
- Enhance Data transmitting: AUTO
- Enhance Addressing: AUTO
- T2 Dispatch: Disabled

Clock Setting Fine Delay
- DRAM CLK Driving Strength: Level 6
- DRAM DATA Driving Strength: Level 8
- Ch1 DLL Default Skew Model: Model 0
- Ch2 DLL Default Skew Model: Model 0

Fine Delay Step Degree: 5ps to 80ps

Ch1 Clock Crossing Setting: AUTO
- DIMM 1 Clock fine delay: Current 2024ps
- DIMM 2 Clock fine delay: Curren 1174ps
- DIMM 2 Control fine delay: Current 1112ps
- DIMM 1 Control fine delay: Current 1112ps
- Ch 1 Command fine delay: Current 74ps

Ch2 Clock Crossing Setting: AUTO
- DIMM 3 Clock fine delay: Current 1900ps
- DIMM 4 Clock fine delay: Current 937ps
- DIMM 4 Control fine delay: Current 886ps
- DIMM 3 Control fine delay: Current 224ps
- Ch 2 Command fine delay: Current 124ps

Ch1Ch2 CommonClock Setting: AUTO

Ch1 RDCAS GNT-Chip Delay: Auto
Ch1 WRCAS GNT-Chip Delay: Auto
Ch1 Command to CS Delay: Auto

Ch2 RDCAS GNT-Chip Delay: Auto
Ch2 WRCAS GNT-Chip Delay: Auto
Ch2 Command to CS Delay: Auto

Common CMD to CS Timing: 1N/2N/AUTO (command rate)

CAS Latency Time (tCL): 5
RAS# to CAS# Delay (tRCD): 5
RAS# Precharge (tRP): 5
Precharge Delay (tRAS): 18
All Precharge to Act: AUTO
REF to ACT Delay (tRFC): AUTO
Performance LVL (Read Delay) (tRD): AUTO

Read delay phase adjust: Enter

Ch1 Read delay phase (4~0)
- Channel 1 Phase 0 Pull-In: AUTO
- Channel 1 Phase 1 Pull-In: AUTO
- Channel 1 Phase 2 Pull-In: AUTO
- Channel 1 Phase 3 Pull-In: AUTO
- Channel 1 Phase 4 Pull-In: AUTO

Ch2 Read delay phase (4~0)
- Channel 2 Phase 0 Pull-In: Auto
- Channel 2 Phase 1 Pull-In: Auto
- Channel 2 Phase 2 Pull-In: Auto
- Channel 2 Phase 3 Pull-In: Auto
- Channel 2 Phase 4 Pull-In: Auto

MCH ODT Latency: AUTO
Write to PRE Delay (tWR): AUTO
Rank Write to Read (tWTR): AUTO
ACT to ACT Delay (tRRD): AUTO
Read to Write Delay (tRDWR): AUTO
Ranks Write to Write (tWRWR): AUTO
Ranks Read to Read (tRDRD): AUTO
Ranks Write to Read (tWRRD): AUTO
Read CAS# Precharge (tRTP): AUTO
ALL PRE to Refresh: AUTO

thx

 

The OCZ memory might be your problem, depending on what speed it is.

I can only assume CPU voltage is Vcore special add, in which case you need to raise it above auto to break free of 3Ghz.

 

well the cpu vid control is the actual thing that changes the voltage...
even if i set the voltage to 1.45v, i still get 3ghz as the absolute maximum
and the ram is running on a divider of 1:2 so its not going over spec which is pc6400 (800mhz)
maybe my cpu is just a bad overclocker?

 

The VID is not what should set the voltage. That should be the identifier for what voltage the chip would normally use when stock.

 

i know what the VID is, but that option does indeed change the voltage...

 

How do you know, have you checked?

 

you were ABSOLUTELY correct sam!
i wad being retarded
thanks for the solution bud!
right now i am at 378 x 9 = 3400mhz but again, it is not stable....
even with 1.4v on the core

 

1.4 is on the low side. Go for 1.45, that's what I used to get my chip to 3.2.

 

what cpu do you have? q6600? if it was, what voltage did it take to get to 3.0ghz?
and if its not too much trouble, could you tell me what batch code your cpu is?

 

I had an E4300, can't remember the batch code, it's currently installed in a system. Would CPUZ tell me? I used 1.45V to get 3.15-3.20Ghz. I used 1.425V to get 3Ghz, but it wasn't very scientific, I tried it and it worked (1.425V) - 1.45V was actually an Autoset.

 

Sam,
A screen shot of your screen should show it. The picture you get if you do the authentication doesn't show the voltage as far as I know.

Russ

 

Well, now the CPU's in my server it no longer has any need to be overclocked:

 

now i am at 320 x 10 = 3200mhz
with 1.36v and it is stable...

 

Sounds about right, your higher end board will probably have less vdroop than my old P35 did back then.

 

i would hope so LOL
anyway, do you still think i can take this cpu higher 24/7stable?

 

1.36V is barely any higher than stock. You could almost certainly go significantly higher with 1.45V.

 

so its not a bad o/c er?