*Official Retail G0 Q6600 Overclocking Thread*

[Semicycle]

Thanks for the feedback all. Pretty satisfied with the core I have now.

[OM DUST]

I have Vid of 1.325 .

Which setup should get me better performance?
I can do 8X425 at 1.425 OR 9 X 400 @ 1.525. My DDR2-800 memory on the 8X425 is a nice o/c at 4-4-412 2T on 2B strap. The system actually feels faster on the 8X425.

[kberguy1]

Thought I'd share my results with everyone. I got this motherboard and cpu from newegg earlier in the week

Q6600 - L743A944 (pack date 12/20/07)
DFI LP LT X38-T2R
Tuniq Tower

Highest I've been able to boot into windows is 3.8Gz, but system freezes and temps idle in the 60s. It also took 1.58v to get this.

I was able to run a tests at 3720Mhz (picture below), but the system isn't stable and load temps reach 75-77c. I believe the CPU Voltage was set around 1.42-1.46 for this test.

I'm currently running at 3600Mhz (1.37v) and all seems stable. Ran a load test last night and the system stayed up without problem. Load temps are hitting 69-73 (depending on core), but the system idles between 48-54 depending on the core. Anyone have any tips for me? I just recently upgraded from my trusty Athlon X2 4200 system I had for about 18months. This is my first Core based cpu system and I haven't mastered everything yet.

Also, Should I be worried about the 4-6 degree difference in the highest to lowest cpu temp readings?

I guess my first comment is that the temps you are getting is inline with what I used to see myself with air cooling. furthermore, that 4-6 C diffrence in core temps is normal. my old B3 stepping Q6600 got within 2 C of each other after lapping, but doing the same with my new G) stepping did nothing. Intel seems to have a bit of an issue with IHS contact with their CPU cores. Having said that, I have a few ideas for you to try.

First, the Tuniq Tower has a removable fan in the middle, and the factory fan is just not as good as some aftermarket units. When I used my Tuniq Tower, I upgraded to a Scythe S-Flex SFF21F, which pushes 60+ cfm without being too loud. it lowered my temps by about 2-3 C, but had an even bigger benefit with PWM temps, which dropped by about 5 C because of extra airflow.

Next, I would focus tuning your GTL levels. Doing so allowed me to lower volts down to 1.275 vCore in BIOS for 3600 MHz, and also lowered my temperatures by about 4 C. Your results may vary, but the right GTL settings make the whole system more efficient.

Third, if you haven't done so, try remounting HSF with a better grade thermal paste (I have used Arctic Silver 5, but am currently using arctic Cooling MX-2). Follow the directions for putting thermal paste on Intel quad core CPU's as outlined on Arctic Silvers website.

[ZenEffect]

i got this pm from a guy who doesnt accept pm's so if you are reading this dude, here you go

Hi,

I noticed in one of the posts you have mentioned that you are willing to share the BIOS and all the settings for Maximus + Q6600 L745A*** batch.

Would you point me to where you have or willing to post those as I have the same mobo and CPU with exact same batch.

I'd really appreciate your help. This will be my forst OC with those quiet expensive toys so, any help is viable.

Thanks,
Mike

my current settings, as im still using this cpu till this weekend, are... (bear with me im doing this all from memory but its accurate)

Extreme Tweaker
Ai Overclock Tuner : Manual
OC From CPU Level Up : AUTO
CPU Ratio Control : Manual
- Ratio CMOS Setting : 9
FSB Frequency : 437
FSB Strap to North Bridge : 400
PCI-E Frequency: 120
DRAM Frequency: DDR2-1165
DRAM Command Rate : AUTO
DRAM Timing Control: AUTO
DRAM Static Read Control: ENABLED
Ai Clock Twister : STRONG
Transaction Booster : ENABLED 0

CPU Voltage : 1.481 (the one right above 1.475 flat)
CPU PLL Voltage : 1.61
North Bridge Voltage : 1.51
DRAM Voltage : 2.26 (my mem rated to 2.35)
FSB Termination Voltage : 1.54
South Bridge Voltage : 1.10
Loadline Calibration : ENABLED
CPU GTL Reference : .63
North Bridge GTL Reference : .67
DDR2 Channel A REF Voltage : AUTO
DDR2 Channel B REF Voltage : AUTO
DDR2 Controller REF Voltage : DDR2_REF
SB 1.5V Voltage : 1.5

NB LED Selection : NB Volt
SB LED Selection : SB Volt
CPU LED Selection : CPU VOlt
Voltiminder LED : DISABLED

CPU Spread Spectrum : Disabled
PCIE Spread Spectrum : Disabled

Advanced CPU Configuration
CPU Ratio Control : Manual
- Ratio CMOS Setting : 9
C1E Suppport : Disabled
CPU TM Function : ENABLED
Vanderpool Technology : Disabled
Execute Disable Bit : Disabled
Max CPUID Value Limit : Disabled


USB Configuration
USB Functions: Enabled
Legacy USB Support : ENABLED

hope this helps, the maximus likes to over volt so watch your cpu ppl, ram, north bridge, and fsb term voltages.

-Effect

[saint-francis]

So these chips have been out for a while now and I'm wondering how they are holding up for people with 24/7 overclocks? I have mine stable with reasonable voltages and temps and my cooling can handle more but I'm only interested in 24/7 settings. Are people seeing any degradation of their performance (needing more voltage than before, lower FSB than before, etc.) with speeds like 3.8 Ghz and higher? I'm assuming I can get to 3.8 with just over 1.4v and keep the load temps still in the mid 50's but I'm not going to do this if it's going to make my system perform worse somewhere down the road.

[mngdew]

L739A466: VID 1.30, needs 1.47-1.48V (after vdroop under load) for 3.6 GHz - average core temp around 65C with Prime95 small FFT test

L744A469: VID 1.325, needs 1.50V (after vdroop under load) for 3.6 GHz - average core temp about 62C-63C with Prime95 small FFT test

Which of these should I keep? The one that needs 0.02-0.03V less vcore or the one that runs 2-3C cooler?

Hmm..hard decision. One needs lower voltage, but the other one runs cooler.
If you are going to o/c not higher than 3.6ghz, keep the one with lower load temp.

By the way, where did you get them from? There was one L744A469 at Tustin Microcenter last week. I picked up an L745A793 as a back up CPU instead of the other, and it looks like I made a good judgement.

[mngdew]

Here is my Load Temps today with 12v on Radiator Fan's (now running DDR2 at 1200Mhz)

You know that there's very little benefit running the memory and FSB async, except on memory benchmarks. It will only stress your memoies and NB, which means more heat.

[The_Barnacle]

My current OC: 3.6Ghz 450x8, 1.475V BIOS, 1.46 idle - 1.42 load
Load temps are around 49-50C
GA-X38-DS4 mother board, 2x2GB Corsair RAM, 5-5-5-15.

Right now all other voltages are at default, but I would like to get it to 475x8 if possible. I think I am FSB limited right now as I can't get 475 to work at lower multipliers either. Any suggestions on what to tweak to try and get 25 more on the fsb?

[Slay0r]

Long story short, my dad got himself a new pc and didn't mind swapping the q6600 i "accidentally found" in it for my pos e4300 (3.2g @ 1.5v).

ram at auto
400x9 1.3v bios set, 1.261 real, 1.264 cpuz (testbed in sig)
hsf = zalman 9500






only had it for about a hour and it's looking good more tests will follow

[Slay0r]

My current OC: 3.6Ghz 450x8, 1.475V BIOS, 1.46 idle - 1.42 load
Load temps are around 49-50C
GA-X38-DS4 mother board, 2x2GB Corsair RAM, 5-5-5-15.

Right now all other voltages are at default, but I would like to get it to 475x8 if possible. I think I am FSB limited right now as I can't get 475 to work at lower multipliers either. Any suggestions on what to tweak to try and get 25 more on the fsb?

try adding 0.1v to vnb.. if that fails 9.0 multi is an option

[PossumK]

By the way, where did you get them from? There was one L744A469 at Tustin Microcenter last week. I picked up an L745A793 as a back up CPU instead of the other, and it looks like I made a good judgement.

That's exactly where I got mine, at the Tustin Microcenter on Saturday. You made a good decision to avoid this one. I didn't really get the option to choose between them since an employee had to bring it out from the back.

I'm going to stick with my original L739A466 with the lower volts. It turns out that the L744A469 needed 0.01V more to be truly stable, and by then the temperatures were about equal. The L744 had a 5C difference between the core pairs while the L739 had only a 2-3C difference. I attribute some of that to the shape of the IHS mating to my lapped Ultima-90 heatsink.

What's interesting is that I see some people with low VID Q6600's, low vcores at 3.6 GHz, and a TRUE 120 getting similar temperatures to my overclock and setup.

[KaOSoFt]

Hello there. I'm here because I need huge support on this.

Here is the information about my Q6600 G0 stepping:

BX80562Q6600SLACR - L737A716 - Packed 10/18/07

This is my VID, and also my idle temperatures using a Thermalright Ultima-90i + Zalman ZM-F3 120mmx25mm fan on the system stated in my signature:



The thing is, I haven't been able to overclock it. Wherever I just up it 60 MHz, it will just restart unexpected. I even overclocked it to 3.6 GHz once, and even passed eight Prime95 hours, and just when I opened Mozilla Firefox to post it on some forums, including this, puff, it restarted, and then it wouldn't even start Windows XP (but it POSTed perfectly).

I've tried all kind of configurations and all. I even lowered the multiplier to 8 (I was using 400x9), leaving the same core voltage in the BIOS (1.4v in BIOS, it would reach 1.296v in CPU-Z at full load), and it happened almost the same: it restarted when I started organizing the files to publish after it successfully passed Prime95 test.

Please, I need some help here. Does any of you has a Gigabyte GA-P35-DS3L motherboard? I assume that's my problem here, but I just bought it, and I've seen good results for other people. I do NOT want to change my board...

Thank you in advance.

[randykahn]

I have the exact q6600 and Gigabyte GA-P35-DS3L system with OCZ DDR2 800mhz memory.
I am very easily able to oc to 3.0 stable:
1. Turn off EIST (in Advanced BIOS screen)
2. Turn off Auto CPU Fan option (so it is always on not whe the cpu gets hot only)
3. go to M.I.T:
- 9X333
-PCI-Express [set this to 100, never set it to default when OC'ing)
- Performance enhance[standard]
-System memory multiplier [2.50v] - check with your memory to see required voltage, Gigabyte's default volt for RAM is quite low
- set CPU voltage to 1.26 [this pretty much depends on your cpu] - what does the "default voltage" shown at the bottom of the MIT screen ?

[KaOSoFt]

I have the exact q6600 and Gigabyte GA-P35-DS3L system with OCZ DDR2 800mhz memory.
I am very easily able to oc to 3.0 stable:
1. Turn off EIST (in Advanced BIOS screen)
2. Turn off Auto CPU Fan option (so it is always on not whe the cpu gets hot only)
3. go to M.I.T:
- 9X333
-PCI-Express [set this to 100, never set it to default when OC'ing)
- Performance enhance[standard]
-System memory multiplier [2.50v] - check with your memory to see required voltage, Gigabyte's default volt for RAM is quite low
- set CPU voltage to 1.26 [this pretty much depends on your cpu] - what does the "default voltage" shown at the bottom of the MIT screen ?

I've already done that.

The default voltage it's the same as my VID: 1.2375v. In fact, I think that what appears there it's the VID.

Thanks for your reply.

[Fixt00l]

IF I am to buy a Q6600 in a couple of weeks, what shall I look for?For now, I only know to look for SLACR for G0, how about the B0 stepping, I am confused.I made a mess at a good forum when asking for the right MB for Q6600+TR ULTRA 120-E, shall it be 8 phase ASUS P5K-E or 6 phase+6 interpolated GB-X38-DQ6...There is a huge price difference here, more than $ 100, and I heard that a Q6600 cannot make more than 8 X 485 on air...

[Shocky]

Next, I would focus tuning your GTL levels. Doing so allowed me to lower volts down to 1.275 vCore in BIOS for 3600 MHz, and also lowered my temperatures by about 4 C. Your results may vary, but the right GTL settings make the whole system more efficient.

[Utnorris]

If you haven't gotten a Q6600 yet and it is going to be a few more weeks, I would just wait for the new Quads to come out. You can already get the Q9300 from Tiger Direct and it should overclock better than a Q6600, IMHO. Of course if you have a Microcenter near you it is hard to justify paying $100 bucks more for the new one when you can get the Q6600 for $200. But if you are going to be paying between $225 - $250 I would just go ahead and get the new one, again, just my opinion.

Utnorris

[theYipster]

How exactly is a Q9300 with a multiplier of 7.5 going to clock better than Q6600 G0 with a multiplier of x9?

[Fixt00l]

Well actually, the Q6600 costs about $ 350 in Bulgaria.And I have read some dissappointing reviews of Q9300.Q6600 has better OC percentage IMHO.

[theYipster]

Exactly.

The Q6600 can be had for as little as $200 at any B&M Microcenter in the US and has better OC potential than perhaps even the Q9450 due its multiplier. Also, let's not forget the degradation issues experienced with the 45nm chips. The Q9300 and the Q9450 will certainly hit modest OCs speeds with great ease. However, to expect a Q9300 to hit 3.6ghz is pushing it, and while the Q9450 should do 3.6ghz on air with a 450mhz FSB, getting one to 4ghz seems to be out of reach in most cases.

On the other hand, seeing a Q6600 G0 hit 4ghz on water, while certainly a great feat, is not uncommon.

[sesion]

Long story short, my dad got himself a new pc and didn't mind swapping the q6600 i "accidentally found" in it for my pos e4300 (3.2g @ 1.5v).

ram at auto
400x9 1.3v bios set, 1.261 real, 1.264 cpuz (testbed in sig)
hsf = zalman 9500






only had it for about a hour and it's looking good more tests will follow

LOL, dude beside showing off some super PI do a few hours prime :>
I can run super pi at 3600 with only 1.22V easy...

[The_Barnacle]

Just picked up another q6600 from microcenter of batch L744B065 package date 1/08/08. Won't have a chance to plug it in and see what the VID or OC potential is on it till next week when my MB arrives.

[Slay0r]

LOL, dude beside showing off some super PI do a few hours prime :>
I can run super pi at 3600 with only 1.22V easy...

Not showing off , (time is crap anyway?) as i said i only had about one hour that day so PI had to cut it.

I haven't been able to test further because under the current heatsink (zalman 9500) temps skyrocket to 80C after about ten seconds of 2*orthos even at 1.26v lol..

I guess the zalman will meet the bin

[highoctane]

So these chips have been out for a while now and I'm wondering how they are holding up for people with 24/7 overclocks? I have mine stable with reasonable voltages and temps and my cooling can handle more but I'm only interested in 24/7 settings. Are people seeing any degradation of their performance (needing more voltage than before, lower FSB than before, etc.) with speeds like 3.8 Ghz and higher? I'm assuming I can get to 3.8 with just over 1.4v and keep the load temps still in the mid 50's but I'm not going to do this if it's going to make my system perform worse somewhere down the road.

I've been running my G0 24/7 at 9x400 1.6vcore for 6 months now, temps max around 65C and it took 1.6 to get over 48hrs quad prime stable.

I may not have gotten a low voltage super clocker, it seems I got one tough vcore loving sob.

[lowtcell]

Q6600 G0 Batch 737B046 Pack Date 11/01/07 VID 1.250

Just started testing my new Q6600 with Xigmatek S1283. Seems to clock easy enough, but I have some concerns on temps.

Idle @ 3600 (9x400) on 1.350v (in Bios)


Prime load went over 80c after a few minutes on CoreTemp so I shut it down.
Idles 33c-37c at stock speed on 1.200v. Gonna try to reseat the HS later and see if temps do better.

Load @ 2400 on 1.200v


Cores @ 51c-57c