*Official Retail G0 Q6600 Overclocking Thread*

[alpha0ne]

L725A

Nice

I have L725A903 turning up early next week

[Funky]

Just received this stepping.
Probably the newest here.
Will be getting Abit IP35 pro soon. Mostly on saturday or sunday.
Fingers crossed.


I havent found anything on this stepping yet. No oc results.

[mascaras]

Just received this stepping.
Probably the newest here.


I havent found anything on this stepping yet. No oc results.

here in PT the steping its even newest (L726A951 ) - TRAY version

[shimmishim]

lol... how many did you get?

[mascaras]

photo its from the Shop Forum , they received them today

[Gnarr]

I got myself a L723A835 SLACR a few minutes ago. It is going to be one hell of a fun weekend

[Pedro Rocha]

here in PT the steping its even newest (L726A951 ) - TRAY version

Mascaras: next time please mention the source of the images.

Please read the XS Forum Policies & Guidelines: http://www.xtremesystems.org/forums/...php?f=120&a=55

Special this one:

26. Plagiarism
We value the content available here, posting content not obtained by yourself without clearly stating so can result in banning.

The source is these images, are the forum of Portuguese store named Aquapc http://www.aquapc-forum.com/showthre...t=67638&page=5


I bought one of these Q6600G0's today on that Portuguese Computer store

Here it is a quick Air Cooled (Tuniq of course) result with a EVGA6860



4X4X4 - 4Superpis X 4 Cores X 4Ghz AR - Easy as 1, 2, 3..4

Nice an unexpensive CPU's indeed

[fornowagain]

Looks like the lower vid value chips are deffinately running hotter at the same vcore as higher vid value chips.

It would, the increase in wattage, hence heat. Is proportional to the change in overclocked voltage squared. A lower stock VID chip at the same Vcore as a higher VID chip has to be hotter.

(1.4/1.3)² = 1.16

(1.4/1.2)² = 1.36

[icon57]

looks like a fair amount of q6600's and evga 680i's are hitting the 4 Ghz mark!

[safan80]

I want to see a lot more 32M superpis at 4.0Ghz please

[Exedy]

I want to see a lot more 32M superpis at 4.0Ghz please

yes..but the temps should sucks ?

I´d like to see a lot more 3.8 32mb at least

[newls1]

L723A825 using P5K-DLX with 1.435V 3.71GHz. I did do a mirror finish lap job on the IHS! I'm really enjoying my Quad Core! Here is a quick picture @ idle right after a 22min SP2004 100%load. Sorry, I forgot to snap a pic while testing for load temps, however when 100% loaded in a 25c room (HOT!!!) the temps were this
core0 = 56
core1 = 56
core2 = 54
core3 = 54

100% loaded!!! This chip OCs better then my L630A E6600 and runs almost cooler. Thanks Intel for the G0 Q6600. I figure that the temps will improve once TIM has cured.

Here is the IDLE temp pics

[arj]

L727A672
7/31/2007

[idiotekniQues]

L727A672
7/31/2007

ouch that is a lot of voltage.

[alpha0ne]

ouch that is a lot of voltage.

1.49Vcore with 1.87Vdimm

[Touay002]

I just got my Q6600 G0 L729A312 Pack Date: 07/31/2007
I removed the E6600 @3600 from The Asus P5B DLX and put in place the Q6600.
running Prime95 now @3600 @1.46V. Temp in the room is ~23C all core are around 61C. So far so good

[obman]

Hi guys,
I am new to OC and have a few questions to the wise and experienced. I am trying to OC my computer for the first time after reading many posts; Some make sense, some I have no clues to what you are talking about. I set it up with the parameters that I found common from many posts as listed below. Please take a look and help me out.

My goal: OC rock solid stable at minimum 3200GHz.
My rig:
Case: Antec 180B
PS: Antec 500W
Processor: Intel Q6600 LGA775 SLACR G0 Stepping Batch 723
CPU Fan: Arctic Freezer Pro 7
MOBO: Abit P35 Pro
MEmory: OCZ 4GB ( 4 x 1GB ) NVIDIA SLI Ready Edition PC2-8500 1066MHz
Cas Latency 5 Timing 5-5-5-15 EPP 5-5-5-15 Volt: 2.1V -2.3V 2.35V EVP**
Video Card: ATI AIWX800XL 256MB GDDR3 2048x1536 resolution
Sound Card: SoundBlaster Audigy2 ZS Platinum
Modem: Sterling Model S20-03B (XP Computer)
HD 320GB Maxtor SATA (Seagate 7200.10) x2
DVD NEC ND-3550A
DVD BQ DW1655
Windows XP Pro

first attempt OC:


Question:
1. Core temp Beta .94 reports temp about 10degrees C greater than uGuru (as seen in the picture). Which one is more accurate?
2. According to Core Temp, baseline is 49C and load is 63C. Is that too high or OK?
3. If that is too high, what can I do to bring the temp down beside switching to better fan like a Tunig or so. Still, Idle temp seems higher than I would expect here.
4. Looking at my parameters, what can I do to improve performance? I just set it in those random numbers and it OC that well (except for the heat). I believe that if I know how to tweek it, I can get so much higher.
5. When I run Prime95, the temps are back down to idle (50C), how is that so? Lower than Orthos? I thought it is a more vigorous test than Orthos? If that is true, my temps are perfectly OK as I am at 57C on all 4 cores. I checked Task MAnager and apparently only 25% of CPUs utilized!! How do I stress it 100%?

All help/pointers are appreciated.

[fallwind]

Orthos will only stress 2 cores, older version of Prime95 will only stress 1 core. D/L Prime95 v25.4 which can launch 4 threads. http://www.computerbase.de/downloads...e95/?url=14026

[obman]

Orthos will only stress 2 cores, older version of Prime95 will only stress 1 core. D/L Prime95 v25.4 which can launch 4 threads. http://www.computerbase.de/downloads...e95/?url=14026

Thank you. Any comments on the rest of the questions?

[fallwind]

-I'd trust the coretemp readings. uGuru is probably reading the temp from the socket area which is always cooler.

-Your temps are OK but run small FFT again with 4 threads of Prime95 and see what happens. If it hits 70c then you probably can't OC any higher.

-to improve performance, you can set the CPU multiplier to 7 and the FSB to 463MHz and adjust the RAM to a 1:1 ratio That'll give you the same CPU speed you have now but much higher bandwidth on the FSB and RAM. Run Prime95 on blend for a few hours to see if it's stable.

-if it pases the previous step you could try a higher RAM divider to speed up the RAM. I believe the next step up would be the 4:5 divider which would put the RAM at DDR2-1158 which might be too high. If it is then you could stick with 1:1 and try tightening the timings, 4-4-4-12 for example.

[Ace-a-Rue]

I just got my Q6600 G0 L729A312 Pack Date: 07/31/2007
I removed the E6600 @3600 from The Asus P5B DLX and put in place the Q6600.
running Prime95 now @3600 @1.46V. Temp in the room is ~23C all core are around 61C. So far so good

that's VERY good overclock at a reasonable vcore!

[ChronoReverse]

Looks like my first G0 can hit 3.6GHz on air after all. Not quite stable though, I'll need to beef up my cooling (unmodified Ninja Plus Rev. B). Should look into lapping the CPU and Ninja.

http://img131.imageshack.us/img131/1...3600mhzji8.png

[dab420]

http://valid.x86-secret.com/show_oc.php?id=228119

[lowdog]

Got two Q6600 G0

1/ FPO/BATCH# L727A619 Pack Date 07/31/2007 with a VID of 1.3000v does 3.6GHz with 1.4v real. Load Prime95 temps are 50-50-48-48 with room ambient of around 25C

2/ FPO/BATCH# L25A903 Pack Date 07/31/2007 with a VID of 1.2650v does 3.6GHZ with 1.4v real. Load Prime95 temps are 55-55-54-54 with ambient temps around 25C


Both do 3.2GHZ with 1.28v real and both are 12 hours prime stable at both frequencies with above settings.

Haven't tried higher frequency with either cpu yet but it appears the lower value VID cpu is running hotter with the same vcore as the higher VID value cpu.

[Ace-a-Rue]

anything over 30 minutes or maybe 60 minutes of priming is overkill...when will you ever run 100% on all four cores...i think the answer is almost "never"!