*Official Retail G0 Q6600 Overclocking Thread*

[jeeka]

Hey ya'll
Currently sitting at 3800Ghz, q6600 g0, 8x475, 1.5v BIOS, 1.48v Speedfan with the gear in Sig, all is well working in "everyday computer workflow" aka no crashed on ANYTHING, but.......

When running 2 instances of Orthos, small FFT's, with the Processor Affinity's set to CPU0/1 for one instance and CPU2/3 for the other, the CPU0/1 instance fails after only 10 seconds at this voltage....the other one just runs and runs (aka about an hour before I stop it)

Is this indicative of not enough vCore? I will check the q6600 OC Database Thread to see what I can generally expect to have to set it to. But would you guys have any general advice?

Please don't waste my and your time posting "up the voltage and find out for yourself jerk"

Your multimedia crunching poster,
jeeka.

[chuckbam]

Hey ya'll
Currently sitting at 3800Ghz, q6600 g0, 8x475, 1.5v BIOS, 1.48v Speedfan with the gear in Sig, all is well working in "everyday computer workflow" aka no crashed on ANYTHING, but.......

When running 2 instances of Orthos, small FFT's, with the Processor Affinity's set to CPU0/1 for one instance and CPU2/3 for the other, the CPU0/1 instance fails after only 10 seconds at this voltage....the other one just runs and runs (aka about an hour before I stop it)

Is this indicative of not enough vCore? I will check the q6600 OC Database Thread to see what I can generally expect to have to set it to. But would you guys have any general advice?

Please don't waste my and your time posting "up the voltage and find out for yourself jerk"

Your multimedia crunching poster,
jeeka.

What is you CPU temps?

[jeeka]

Temps are outstanding for AIR and this time of year:

At this OC,
Idle:
Core #0 = 39C
Core #1 = 37C
Core #2 = 32C
Core #3 = 35C

Orthos (load):
Core #0 = 62C
Core #1 = 62C
Core #2 = 55C
Core #3 = 55C

I am aware of the disparity between core temperatures, and mine are slightly out of the norm of what everyone else is measuring. I can chalk it up to:
1) That's the way it is, deal with it!
or
2) I had a slightly slanted (not concave or convex) Thermalright 120 Ultra Extreme that might be putting unequal pressure on the chip when screwing in the mounting screws with equal pressure. I attempted to fix it by lapping down the "top of the hill" side of the slant as much as possible and doing the "Washer between the mounting bracket and HS base trick" that you guys hooked me up with. That actually dropped temps about 3C.

[chuckbam]

Temps are outstanding for AIR and this time of year:

At this OC,
Idle:
Core #0 = 39C
Core #1 = 37C
Core #2 = 32C
Core #3 = 35C

Orthos (load):
Core #0 = 62C
Core #1 = 62C
Core #2 = 55C
Core #3 = 55C

I am aware of the disparity between core temperatures, and mine are slightly out of the norm of what everyone else is measuring. I can chalk it up to:
1) That's the way it is, deal with it!
or
2) I had a slightly slanted (not concave or convex) Thermalright 120 Ultra Extreme that might be putting unequal pressure on the chip when screwing in the mounting screws with equal pressure. I attempted to fix it by lapping down the "top of the hill" side of the slant as much as possible and doing the "Washer between the mounting bracket and HS base trick" that you guys hooked me up with. That actually dropped temps about 3C.

That is good, since the Thermal Spec is 71c. You have some room there.
In addition, you have that temp with the high Vcore.

Have you attempted or considered a stress test like, Orthos and 3Dmark06 at the same time. It would be good if you can complete 3Dmark06 under that configuration.

[jeeka]

That is good, since the Thermal Spec is 71c. You have some room there.
In addition, you have that temp with the high Vcore.

Have you attempted or considered a stress test like, Orthos and 3Dmark06 at the same time. It would be good if you can complete 3Dmark06 under that configuration.

Good idea....Should I try 2 instances of Orthos AND the 3D Mark 06 run? or 1 Orthos + 3D Mark 06?

Also, should I assign affinities to each program or just bombard the chip and let it sort it all out?

[chuckbam]

Good idea....Should I try 2 instances of Orthos AND the 3D Mark 06 run? or 1 Orthos + 3D Mark 06?

Also, should I assign affinities to each program or just bombard the chip and let it sort it all out?

I am testing now, It looks like I will end up with 3.27Ghz, (9x 364) 1456 FSB
I have been looping 3DMARK06 and Prime95 ver. 25.4 (because it is a 4 core test out of the box). I had many BSD's today. So, I think it is a good test.

The max I can go is 9x 366. So I pulled it back a little. I am running 1.3 volts under-load and 1.33v at idle. MAX temp was under extreme load is 68.c, but mostly under hi load around 64c. Idle, in the mid 40c. New heat sink compound today, so maybe that may get a little lower in time.

With 9x 366 I looped the 2 programs for about 3 hours with no errors today.
I just came back home happy to see all was good. Going to run some benchmarks, boot into Vista and run Prime for the night.

On my every day use, I use Power management settings that drop me to a 6x multiplier at idle. Vista performance setting, Speed Step and C1 (EST?) enabled, Power Fan options turned of for max 24/7 air flow.