Sign in with
Sign up | Sign in
Your question

Help overclocking Q6600 to 3.8+ with lower voltage

Last response: in Overclocking
Share
May 17, 2008 2:20:48 PM

Hey all, thankyou for taking interest in my thread, I am looking for some general help and tips on how to overclock my Q6600 to 3.8GHz or higher for benching at a lower voltage.

I will start by listing my hardware:

Gigabyte X38-DS5 DDR2 Motherboard

8 GB OCZ Reaper @ 1035MHz 5-5-5-15 (2.35v)

OCZ GameXstream 700w PSU

250GB, 320GB Seagate Hard Drives

Q6600 @ 3.72GHz 1.325 VID (1.7v) LOL! Temps: 34,34,35,36! - Weird!

ATI Radeon 3870X2 @ stock

Sound card etc.

Ok, what I am obviously concerned about is my crazy voltage on the CPU (1.7v) and the fact that my temps are so low! I have left all other voltage settings on Normal as they don't seem to make any effect when using lower voltage, but I will list the options in my BIOS available and what I have them on:

CPU Voltage: 1.7v Frequency= 3.72GHz 9.0x Multiplier

FSB= 414MHz 9.0x

PCI-E Frequency: Auto

C.I.A.2: Disabled

Performance Enhance: Standard

RAM: 1035 5-5-5-15 2.35v - (voltage too high) - My motherboard has a habbit of automatically lowering the default voltage to 1.8v from 2.2v

Advanced RAM timings: Auto

DDR2 Voltage: 2.35v

PCI-E Voltage: Normal

FSB Overvoltage: Normal

MCH Overvoltage conrol: Normal

MCH Referance voltage: Normal

DDR Referance voltage: Normal

DDR2 Termination: Normal

Load Line Calibration: Enabled!

VID: 1.325V

Idle temps: 33-37c on all 4 cores.

- These are all the BIOS settings I have for tweaking etc, there may be an advanced menu? But I would really like some help at getting my CPU to 3.8GHz+ at a much lower voltage if I can, or is it just a bad CPU? I have not Orthosed the CPU at these settings yet but it seems to be stable in Games and Benchmarks etc, thankyou for your help! - Alex.
May 17, 2008 5:52:08 PM

Anyone? I am using voltages higher than that of what a B3 requires lol, is it my board? Or am I just not doing it right Thanks.
May 17, 2008 6:08:52 PM

Sounds that you are too right.LOL.
Sound damn nice to have those temps at such high voltage.
I might left out something but everything seems good.
Related resources
May 17, 2008 6:40:59 PM

Humm, so am I okay too but high voltages through the chip as long as I have good temps? Thanks.
May 17, 2008 7:08:39 PM

TehPwn said:
Anyone? I am using voltages higher than that of what a B3 requires lol, is it my board? Or am I just not doing it right Thanks.

Sounds wierd
Which software you are using to ckeck your temps??
What is your ambient (Room) temp ??
May 17, 2008 7:48:44 PM

Its your Chip, not stepping. You have the Worse VID possible for a q6x00 series. That means your Chip starts as high as possible for a g0.

As you near upper voltages and speed, your chip will automatically need more VCore to be stable, because the VCore becomes less efficient as you raise it.

For instance my G0 1.3250 VID chip needs 1.440 Loaded to opporate at 3.6. (Modifying the GTL refs and junk relinquished the same speed at 1.420 VCore loaded.)

So, besides for high voltage anyways, you'll need a new chip to change that! Damn intel stripping their low VIDs from retail boxes! You have a much better chance at a OEM chip, because its possible that the OEM bought 100,000 a year ago or more, before intel was hip to the VID mattering to us OCers.

--Lupi
May 17, 2008 8:02:24 PM

Humm, with my 1.325 Lupi I need 1.6-1.65v just for 3.65GHz! Now surely thats not right, as my temps are so low I do not mind going for high voltages, what should I increase in order to achieve 3.8ish with a voltage lower or around 1.7v lol Thanks.
May 17, 2008 8:20:04 PM

Already know the values of your VDrop and VDroop? If not, boot at factory VCore setting, (The VID) Then load windows and note CPUz's idle violtage. The difference will reflect your initial VDrop. Then test with Prime 95 small ffts for 1 minute, and note the now lowered VCore voltage in CPUz.

Once you get to know those two, you can directly target a VCore voltage without having to look! Also, does your mother board support any version of Load Line Calibration?? That needs to be on big time when testing that high.

--Lupi
May 17, 2008 8:37:51 PM

Yes it suppoprts Load Line Calibration thankfully, the Vdroop is not much about -0.10 or something, and whats VDrop? Humm, could it be my board at fault? Lupi whats the highest voltage I could put through my chip without it dying on me in a monthor so? Thanks.
May 17, 2008 8:58:29 PM

At 1.7v, you're risking instant burnout regardless of temperature. It's not just a matter of lower lifespan, the chip can go at any minute. Tune it back down. :na: 

It's strange you'd need such high vcore just to reach 3.72ghz. My VID 1.3125 q6600, which is almost as bad, run at 1.65v for 4ghz, one hour prime95 stable. Didn't test for more than 1 hour, 1.65v is too hardcore for me. :p 
May 17, 2008 9:10:42 PM

hmmm... try with just 2 sticks of ram... I think it might overclock a little better..
May 17, 2008 9:11:03 PM

Ahhh! LOL! All this is annoying me now! Ha OMG Dagger you have nearly the same motherboard as me (mine is the DS5 version) So what settings did you use in BIOS to get to these settings? Ha maybe we can get somewhere with this?!? Thanks.
a b à CPUs
a b K Overclocking
May 17, 2008 9:17:32 PM

@OP: You should set the PCIe to 100.
a b à CPUs
a b K Overclocking
May 17, 2008 9:18:30 PM

Also try the following:

Increase MCH voltage a little
Increase FSB voltage +.1
May 17, 2008 9:19:37 PM

TehPwn said:
Ahhh! LOL! All this is annoying me now! Ha OMG Dagger you have nearly the same motherboard as me (mine is the DS5 version) So what settings did you use in BIOS to get to these settings? Ha maybe we can get somewhere with this?!? Thanks.



The only thing that's off is PCIE clock. You didn't lock it at 100mhz. Overloaded graphics card can cause loss in stability.

The ram voltage is higher than I'd think is safe, but up to you. :p 

I had no problem running it at stock chipset voltage, but it doesn't hurt to try higher.
May 17, 2008 9:20:35 PM

Humm, MCH doesnt seem to have any effect with me LOL! By how much? I have put as much as + 2.25 through but still no luck LOL! Could my board be faulty?
May 17, 2008 9:21:59 PM

TehPwn said:
Humm, MCH doesnt seem to have any effect with me LOL! By how much? I have put as much as + 2.25 through but still no luck LOL! Could my board be faulty?

Then it's not the chipset voltage. There is no way x38 requires such high voltage for that tiny overclock. Remember, it runs at 1600mhz fsb stock.
May 17, 2008 9:45:06 PM

So you left everything on Auto except PCI-E Frequency? I just set that to 100 with FSB Voltage increased by .1+ and MCH increased a little and it would even boot into Windows with a 1.65 voltage, IMO what next? Should I get a new board? Thanks.
May 17, 2008 9:45:50 PM

Just a thought, could my Power supply be limiting my Overclock in any way? Its a 700w OCZ, thanks.
May 17, 2008 11:38:35 PM

Anyone???
May 17, 2008 11:45:03 PM

I use a 700 watt tough power on my 3.8 OC, so prolly not. Its more than likly the VID of your chip. Its just to high, because there is a sliding scale at work. From 3.6 to 3.8 takes an exceptional amount of VCore to stay prime stable. As a matter of fact, thats the first q6600 I have gotten 8 hour stable, and its a 1.2500. So I get like 16 extra VCore notches you dont, and that alone makes it perform better. The more VCore you add, the more it needs.

And then at 3.5 on a 1.3250 you encounter the High VID syndrome where it takes an exceptionally large ratio of VCore to speed gains where it doesnt make it worth it. If it starts high, it will need to go much higher. Odd, but thats how it works.

I only know it because I tested 10+ q6x00s myself.

--Lupi
May 17, 2008 11:47:40 PM

Oh, I notice that at 1.2625 the syndrome is minimal up to 3.6. My 1.2750 one reflects the high VID syndrome right before 3.6. My 1.2500 that I tested at the 3.8 showed it at like 3650.. so that means my 1.200 shouldnt encounter it until around 3.7ish. So I may get 3.9 on it!


Weeeeee!

--Lupi
May 18, 2008 12:07:55 AM

Humm interesting VID stuff! Thanks Lupi! But hey evenm Dagger said he had a 1.3125 VID and only needed 1.65v for 4.0GHz for a quick bench etc! I need 1.65v for just 3.65GHz, now surely his VID cant be that much better than my 1.325v one, his is the next worse one! So what should i do? LOL
May 18, 2008 12:42:24 AM

Oh, I dont mean for a bench. I meant to actually be prime stable for hours. I could prolly fake a bench at 4.0 or so at the same VCore value.

--Lupi
May 18, 2008 12:58:46 AM

LOL! So is it not impossible to get 4GHz on a 1.325v VID or is it? I mean I don't mind going to around 1.6-1.7v if needed, but I just need to know if a 1.325 VID would completely stop me from getting 3.8GHz or above??? Im starting to think the problem is more motherboard related??? Thanks for your help so far! - Alex.
May 18, 2008 3:12:40 AM

TehPwn said:
LOL! So is it not impossible to get 4GHz on a 1.325v VID or is it? I mean I don't mind going to around 1.6-1.7v if needed, but I just need to know if a 1.325 VID would completely stop me from getting 3.8GHz or above??? Im starting to think the problem is more motherboard related??? Thanks for your help so far! - Alex.

Lol, it's not like that extra speed really matters anyway. :p 

But 1.65vcore just for 3.65ghz is way too high. I use 1.45v for 3.6ghz, 5+ hours prime95 stable (didn't test more), on a VID that's almost as bad as yours. There seems to be a real wall past 3.6ghz though, as 3.8 takes at least 1.6v to be stable. What vcore does it take for an even 3.6ghz (1600mhz fsb)? Don't go above 3.6 if there's a wall there.
May 18, 2008 3:58:52 AM

LOL! With our vids being so close, I expect our expectations are similar! I benchmark you see, so I would just like 4GHz or so at whatever voltage for just a quick bench (no long term stability lol) so im guessing it is rather motherboard related as you need 1.45v for 3.6 so In reality I would need that or 1475v or so, thanks for the help, Alex.
May 18, 2008 4:32:07 AM

dagger said:
Lol, it's not like that extra speed really matters anyway.
Truth
May 18, 2008 5:02:40 AM

The only thing I can tell you is I have had a b3 at 1.8250 in the bios, 1.77 loaded. It didnt harm it, just made the cores 80c! So... it will hurt, but who knows how much? if you wanna get another chip anyways, I guess you can find out! I think it will be somewhat maybe possible to get 3.8 on that chip. But I am not sure!

Just be careful what you do, unless it doesnt matter to much!

--Lupi
May 18, 2008 5:27:55 AM

My q6600 runs at 3.4 ghz at 1.28v. Your 1.7 seems insane to me lol. I'm surprised It'll go that high.
May 18, 2008 5:46:11 AM

TehPwn said:
Humm, so am I okay too but high voltages through the chip as long as I have good temps? Thanks.
You should read this link on Electromigration.
May 18, 2008 1:14:19 PM

Hey thanks for all your support, bobert866, whats is your VID? Cause mine is 1.325v and Daggers is 1.3125v which is more or less the same and he can get 4GHz with 1.65v and I cant even get 3.7GHz with 1.7v LOL! Lupi will a 1.3125 VID and a 1.325 VID make the difference, will all 1.325 VID's just lock up and require a load of voltage to get anywhere? Cause me and Dagger have the same BIOS and even when I use his settings I get nowhere lol, do you think it might be graphics card related as I have a 3870X2 without the 8 pin adapter, could this be causing power problems? I suspect its now a motherboard problem, but hopefully we can get somewhere lol, thankyou all for your support, Alex.
May 18, 2008 1:42:47 PM

TehPwn said:
Hey thanks for all your support, bobert866, whats is your VID? Cause mine is 1.325v and Daggers is 1.3125v which is more or less the same and he can get 4GHz with 1.65v and I cant even get 3.7GHz with 1.7v LOL! Lupi will a 1.3125 VID and a 1.325 VID make the difference, will all 1.325 VID's just lock up and require a load of voltage to get anywhere? Cause me and Dagger have the same BIOS and even when I use his settings I get nowhere lol, do you think it might be graphics card related as I have a 3870X2 without the 8 pin adapter, could this be causing power problems? I suspect its now a motherboard problem, but hopefully we can get somewhere lol, thankyou all for your support, Alex.

Speaking of graphics card, did you lock PCIE at 100mhz yet? If not, it might be the problem. Graphics cards don't like bus overclocking.

And try setting the ram back to 1:1 ratio, which is 800mhz on 1600mhz fsb (3.6ghz). Those ram voltage are high too. Don't push those things too hard. :p 
May 18, 2008 1:55:38 PM

Yea, PCI-E is set to 100MHz with a slight overvolt on the PCI-E too, and I am using a 3:4 ratio, I had no look with 1:1, my RAM is now at 1150MHz 5-5-5-18 and im in Windows at 3.8GHz for about 1 hour now no problems so far lol at 1.7v, ahh, this has gotta be a Motherboard problem, surely VID doesn't matter that much does it lol?
May 18, 2008 2:07:15 PM

TehPwn said:
Yea, PCI-E is set to 100MHz with a slight overvolt on the PCI-E too, and I am using a 3:4 ratio, I had no look with 1:1, my RAM is now at 1150MHz 5-5-5-18 and im in Windows at 3.8GHz for about 1 hour now no problems so far lol at 1.7v, ahh, this has gotta be a Motherboard problem, surely VID doesn't matter that much does it lol?

PCIE should not be overvolted. It won't help, can only create problems. :p 

You can also try to further loosen timing to 6-6-6-18 and see if it helps.

Also, what temperature is it running at on full load? At 1.7vcore, you should definately keep an eye on that. In any case, 1.7v isn't what's usually considered safe. You shouldn't run it for an extended period of time.
May 18, 2008 2:07:58 PM

What I can't believe is that no one has called this BS yet. You need to double and triple check those temperatures. Hopefully, you're phase change or high-end water cooling or those temps are next to impossible. If you say you're air cooling, then you're definitely seeing the wrong temps and that chip is going to go bye bye in your near future.

The really seasoned vets on this board with the higher bin chips won't run 3.8 for long, and thermals were their limiting factor. It's fun to push it, but if you plan on keeping it, back it down and properly calibrate your thermals.
May 18, 2008 2:47:05 PM

Actually those temps are correct! Ive checked with Core Temp, HW Monitor and Everest and they ALL report the same! Actually I was playing FEAR for 1 hour yesterday and the temps never went above 47c! Amazing I know, and nah I don't really intend to keep it that long, ill invest in nehalem when it comes out or buy a penryn when they become cheaper, so I guess id expect 1 year or so and thats about it, im thinking of getting a new motherboard though.
May 18, 2008 3:21:45 PM

u need to run prime95 or OCCT and then tell me your temps. Idle temps mean nothing.
May 18, 2008 4:16:50 PM

Ok, so would you people (prefferably Lupiron - man I love that guy!) suggest that its the fact that 1.325 VID's just simply cannot hit 3.8GHz+ without crazy voltage or is my problem more motherboard related??? Thanks.
May 18, 2008 5:29:01 PM

Why does it have to be Lupi? :fou: 

:kaola: 

Lupi did some research with the relationship with VID and OCing. Basically, what he concluded was that the higher the VID, the worse the OC, or rather the OC requires more voltage. Jumping from 3.6GHz to 3.8GHz requires exponential amount of voltage, so high voltage is expected for all Q6600s. Normally 3.8GHz requires something like 1.6-1.65V, and considering you have a high VID Q6600, it's reasonable once the Vdroop and Vdrop is factored in.

As for the temps, what is your room temps? Do you have speedstep enabled?
May 18, 2008 5:38:08 PM

TehPwn said:
Ok, so would you people (prefferably Lupiron - man I love that guy!) suggest that its the fact that 1.325 VID's just simply cannot hit 3.8GHz+ without crazy voltage or is my problem more motherboard related??? Thanks.



It's you. You're not doing it right. But don't worry, first time hurts, that's why you keep doing it. Or you can have someone older and more experienced teach you. :sol: 

Edit: Err... what was I just thinking about? Oh yeah, post all your bios settings, so we can see what's wrong. :sarcastic: 
May 18, 2008 5:42:51 PM

Ha! Well no speedstep is disabled, my room temps I would say is around 20c, maybe lower, I have a window open and no heating on, so yea :)  But even Dagger with your 1.3125 VID you aid you got a quick 4GHz with 1.65v, now surely that would mean that I wouldn't need too much over that either just for a bench? I hope we can get some headway anyway lol, thanks so far :)  - Alex.
May 18, 2008 8:47:27 PM

Anyone??? I really need advice, should I just give up on this motherboard and get the ASUS Rampage Formula X48, or is it the CPU I should give up on? Or both? LOL! Many thanks - Alex.
May 18, 2008 8:57:52 PM

Its the CPU! Like I said before, it takes simply a huge amount of VCore voltage to become stable! As in, ALOT more than you'd expect!

A Rampage is an excellent board, since I used a formula maximus to get the results I did from my 1.2500 chip.

I would guess you need a better VID chip. Unless you have such a huge VDrop and droop where that board nullifies the gains you want!

Did you ever measure the VDrop and VDroop? Thats a large factor as well!

--Lupi!
!