overclocking q6600 g0 and abit ip35pro

[robr]

I just put together a new machine and I'm now trying to OC it. I've followed the OC guide in the sticky but I'm running into problems with the simplest changes.

I have an Abit IP35 Pro w/Q6600 G0
4x1 gig of Crucial PC6400 (CPU-Z shows 5-5-5-18, but this is 4-4-4-12 memory per NewEgg - link here http://www.newegg.com/Product/...0146565&Tpk=20-146-565
)
Tuniq Tower 120

as well as a 620HX Corsair power supply
8800GTS superclocked video

So the first thing I did was try to use the abit uGuru utility to overclock. This utility does not allow you to select auto voltages, but auto voltages was set in the BIOS, so I'm not sure which applies.

Default is 266FSBx9

From what I've read, I should have no problem getting to around a 380x9, however again using uGuru, the best I've done is a 320x9@1.375v (I did not play with voltage to see how low I could get because 320 is simply too low, that was just an arbitrary number I picked).

360x9@1.375 blue screened, so I decided to lower the FSB and up the voltage
349x9@1.405 hung immediately upon hitting APPLY

 

[robr]

I just changed the memory settings from Auto to 4-4-4-12 and now can't POST. The board LED shows a C1 which is 'Detect memory - auto detection of DRAM size, type and ECC'

 

[swtethan]

non bios overclocking is yucky

 

[robr]

It appears it is BIOS overclocking, just with a windows based app controlling the BIOS settings. ie, if i make a change in uGuru, then reboot and go into BIOS, the change has been made in BIOS. In any event, I've made it a little bit further.
I reset the BIOS, then tried with x8 this time. I was able to boot into windows at 425x8 (all settings in BIOS are automatic except the 425x8). uGuru reported CPU voltage as 1.25 (but again this was on an auto setting). As soon as I fired up Orthos though, the PC hung within 2 or 3 seconds.

 

[nyker96]

you can test with non bios OC but use BIOS oc for permanent settings.

 

[robr]

I've made some progress. The issues appear to have been with my memory settings, default was 1.8v while this memory required 2.2v. I manually changed DDR voltage to 2.2, then set the memory to 4-4-4-12.
I hard set CPU voltage to 1.5 and a multiplier of 8. Then I started using a combo of uGuru to change the FSB speed, while using Everest to stress the CPU/memory.

I was able to get to 451x8 with a steady temp of 51-53C while running Everest for about 5 minutes. 460x8 hung as soon as I started Everest. I've set it back to 451 and kicked it off again and will let it run either until I figure out it appears solid and temps are sticking, or I find that's too much voltage to be running (which is what I'm off to read about now).

My memory is set to 1:1, is that what I want? During POST it says DDR2-902

 

[robr]

I just took a different approach. 389x9@1.325v. It's slower than I was able to achieve with x8, but it's also 43C instead of 51-53C.

 

[Shimmishim]

Originally posted by: swtethan
non bios overclocking is yucky

do you own this board?

actually uguru is linked directly to the bios settings.

if you change voltage in uguru, it changes it in bios.

it's something that's been on abit boards for a while and works marvelously...

i.e. if you're stability testing and orthos fails due to low voltage.. you don't have to reboot to increase your vcore... you just open up uguru, change your vcore and run again....

when you actually go to reboot, your news settings have been set and you can confirm this in bios...

same thing with FSB and other voltages as well... i don't think any other company's board can do that...

 

[robr]

Well, I'm admitting I'm a failure. After running Prime95 with the above settings, I found the system to be unstable. I'm back to 405x8@1.375v and it's stable, but I was hoping it would be fairly simple to get to 3.5GHz with a G0. Nobody else seems to have much issue, and I know my gear shouldn't have a problem, so I can only assume the problem is me .

 

[Shimmishim]

what are your mch, vtt, ich, and ichio voltages?

 

[robr]

I haven't touched those (since I don't know what they are and don't want to blow anything up).

They are:

MCH 1.25
VTT 1.20
ICH 1.05
ICHIO 1.50

 

[robr]

After reading a little about the above settings, I'm now running

391x9@1.425v
MCH 1.33
VTT 1.23
ICH 1.15
ICH10 1.60

DDR2 2.2V (memory rated at 2.2V)
4-4-4-12
1:1

Core temps are 67-69 under load with PRIME95
Room temp is about 78F

Still too much voltage and too much heat for my liking, especially after seeing people hit these speeds with 1.2xV

 

[ChronoReverse]

I doubt anyone is hitting 3.6GHz on air without raising voltages on retail G0 Q6600's...


Getting 3.5+GHz on dual cores may not be unusual for G0's but quad cores are a different story.

 

[BonzaiDuck]

When it comes to over-clocking, I'm the equivalent of a political right-wing. I'd rather drop the CPU_FSB and tighten the latencies to minimum. Or drop the multiplier and see how much I can pump up the CPU_FSB without loosening the latencies much.

I can see you have this thing pumped up to 3.5 Ghz, 1:1, and several people are doing this with the G0 stepping.

The temperatures you're getting -- if your room-ambient is accurately measured -- would still seem about right, but you haven't said much about your cooling solution. Are you using the stock Intel HSF?

Others may comment -- per the MCH (assumed 1.2VHT), CPU_VTT voltages, and even your chipset voltages -- I thought I noticed that edging these things up a tad would allow stable over-clocks at lower VCORE -- and therefore lower temperature (slightly). Could be my imagination, though.

The VTT voltage is your CPU's FSB voltage. As you raise your FSB speed, increasing the VTT a tad will help getting to a stable over-clock.

 

[robr]

Bonzai, thanks I'll keep tweaking. I did mention my cooling in the very first post in this thread. I'm using a Tuniq Tower 120. My goal here is 3.5GHz with reasonable voltage and temps, I'm not looking to squeeze every last bit out of this. I just want to be cool and safe. I thought I was stable at 380x9@1.375V (3375) but PRIME95 failed about 90 minutes in. It was my hope to be able to run a stable 3.5GHz at under 1.4V, but I'm just not getting there. I think I'm having more success at x8.

 

[Blacklash]

Definitely try working from 8x. Fellows @ XS seem to be doing well that way.

I'll be doing the same when I switch to x38 or P35. Currently, I am on a BX2 and it will not let me above 390 x 9. So it's 3.5 @ 1.45v for me, for now.

My system right now is as below-

Intel Core 2 Quad Q6600 @ 3.51GHz 1.45v (24/7)|Mushkin XP2-6400 (4-4-3-4)
Intel D975XBX2 (i975x)|EVGA GeForce 8800 GTX (659|2040)|NEC 20WMGX2-BK
WD Raptor 74GB HD(x2)|Pioneer-710B5PK|NEC ND-3550A|Klipsch ProMedia Ultra 2.0
SeaSonic S12 EP SS-650HT|Antec P180 B|TR U-120 Extreme|Vista Home Premium x64

 

[bryanW1995]

Originally posted by: robr
After reading a little about the above settings, I'm now running

391x9@1.425v
MCH 1.33
VTT 1.23
ICH 1.15
ICH10 1.60

DDR2 2.2V (memory rated at 2.2V)
4-4-4-12
1:1

Core temps are 67-69 under load with PRIME95
Room temp is about 78F

Still too much voltage and too much heat for my liking, especially after seeing people hit these speeds with 1.2xV

your voltages are fine. I'm at 1.5 just to get 8x445 on an e6750, also using a tuniq.

 

[bryanW1995]

Originally posted by: Shimmishim

Originally posted by: swtethan
non bios overclocking is yucky

do you own this board?

actually uguru is linked directly to the bios settings.

if you change voltage in uguru, it changes it in bios.

it's something that's been on abit boards for a while and works marvelously...

i.e. if you're stability testing and orthos fails due to low voltage.. you don't have to reboot to increase your vcore... you just open up uguru, change your vcore and run again....

when you actually go to reboot, your news settings have been set and you can confirm this in bios...

same thing with FSB and other voltages as well... i don't think any other company's board can do that...

why didn't I read this before all of my earlier stability testing??? Great info! I disabled uguru after it locked me up while trying to alter the TT speed when it was already set on manual mode, so I never figured this out.

 

[cmdrdredd]

Originally posted by: bryanW1995

Originally posted by: Shimmishim

Originally posted by: swtethan
non bios overclocking is yucky

do you own this board?

actually uguru is linked directly to the bios settings.

if you change voltage in uguru, it changes it in bios.

it's something that's been on abit boards for a while and works marvelously...

i.e. if you're stability testing and orthos fails due to low voltage.. you don't have to reboot to increase your vcore... you just open up uguru, change your vcore and run again....

when you actually go to reboot, your news settings have been set and you can confirm this in bios...

same thing with FSB and other voltages as well... i don't think any other company's board can do that...

why didn't I read this before all of my earlier stability testing??? Great info! I disabled uguru after it locked me up while trying to alter the TT speed when it was already set on manual mode, so I never figured this out.

The only thing it cannot change is memory timings and ratio.

 

[bryanW1995]

yeah, I had to reset cmos earlier b/c I couldn't get the stupid thing to boot at 8x441. It worked fine after I booted at 8x439 and changed it in uGuru, however.