Where to go now?

H

Hades16x

2[H]4U
2FA
Joined
Sep 24, 2003
Messages
3,976
When you guys shoot for maximum stable overclock, how do you achieve it?

I would assume you first jump major gaps, like to the minimum overclock you've seen on the processor.

Let me be more specific.

I sold my E6600/evga680i/8800gtx to my best friend for 300 bucks, he will be receiving it once the 9800 series comes out, as I will be going to a x38/e8400/ and some form of the 9800 series.

He is not very good on the overclocking front, so I'm trying to achieve maximum on mine/his E6600 and 8800gtx.

Using google as my friend and the C2D overclocking faq, I arbitrarily set my Vcore at 1.45 at set my FSB to 1425, 3206mhz core.

Then I ran Orthos for a simple 20 minutes without any failure and temperatures stayed below 53C via SpeedFan and RivaTuner with C2D plugin.

Now I know the only way we can achieve full stability, or stability as long as we need it for, is to run orthos for that full length of time without error.

So I reboot, set FSB to 1466, 3.3ghz even effective core. Start orthos, it fails in between 10-13s. Idle temperatures are 30/31C, core1/core2, which doesnt tell much.

CPUz states core voltage at 1.416.

Now, I have no BSOD, and I'm currently typing this post at 3.3ghz and it has been running for nearly 30 minutes stable at this speed.

What should I do? I haven't bumped up the Northbridge voltage from 1.4 to 1.5. Can I do this on a stock 680i northbridge cooler without ill effect? I know on x38/p35 board with e8400 you usually have to up NB voltage do to over 4ghz on water.

The ram is running at 2.150V at 400mhz, so I'm not overclocking the ram yet.

Should I back down to 1425, and increase 5fsb at a time, until I get a failure on a 20 minute orthos, then back down 5fsb and see if its 24 hour stable?

Maybe I'm expecting too much, but I feel that if I'm windows/idle stable, that I can get this 100% stable.

Or do I bump up Vcore?

Thanks,
Michael
 
Type: Blend - stress CPU and RAM Min: 8 Max: 4096 InPlace: No Mem: 3836 Time: 15
CPU: 3300MHz FSB: 366MHz [366MHz x 9.0 est.]
2/29/2008 9:43 PM
Launching 2 threads...
1:Using CPU #0
2:Using CPU #1
2:Beginning a continuous self-test to check your computer.
2:press Stop to end this test.
1:Beginning a continuous self-test to check your computer.
1:press Stop to end this test.
2:Test 1, 4000 Lucas-Lehmer iterations of M19922945 using 1024K FFT length.
1:Test 1, 4000 Lucas-Lehmer iterations of M19922945 using 1024K FFT length.
Type: Blend - stress CPU and RAM Min: 8 Max: 4096 InPlace: No Mem: 3836 Time: 15
CPU: 3300MHz FSB: 366MHz [366MHz x 9.0 est.]
2/29/2008 10:11 PM
Launching 2 threads...
1:Using CPU #0
2:Using CPU #1
1:Beginning a continuous self-test to check your computer.
1:press Stop to end this test.
2:Beginning a continuous self-test to check your computer.
2:press Stop to end this test.
1:Test 1, 4000 Lucas-Lehmer iterations of M19922945 using 1024K FFT length.
2:Test 1, 4000 Lucas-Lehmer iterations of M19922945 using 1024K FFT length.
1:FATAL ERROR: Rounding was 0.5, expected less than 0.4
1:Hardware failure detected, consult stress.txt file.
1:Torture Test ran 0 minutes 21 seconds - 1 errors, 0 warnings.
1:Execution halted.

2:Torture Test ran 0 minutes 22 seconds - 0 errors, 0 warnings.
2:Execution halted.
 
Okay next step I took.

Rebooted and went into bios to look at Vcore. Now I understand vdroop. I had it set at 1.450. the value output stated was 1.40.

Now if we assume a properly cooled E6600 can handle 1.50v Vcore, does that mean with vdroop I will have to set vcore to 1.550?

I set FSB to 1435 and vcore to 1.475V. Going to run Orthos now.

Thanks,
Michael
 
1.4-1.45v should be plenty for a Q6600 at the speeds you're trying to run it at. It seems like you're limited by your chipset and not your CPU, so try raising the MCP voltage one step at a time and see where that gets you.
 
20 minutes stable at 1435.

Can anyone verify if vdroop is linear or exponential?

If we assume my motherboard has a .05V droop, will my e6600 be fine at 1.550v?

Thanks,
Michael
 
Man this is so fun.

vdroop is not linear, at least not with the accuracy the mobo will give me.

at 1.45v I got 1.42v via CPUz
at 1.475 I got 1.43v via CPUz
at 1.48125 I get 1.44v via CPUz

my first trial at 366fsb at 1.45v failed orthos at under 20s.
at 1.475 it failed at the 4 minute mark.

I think I'm closing in on the stability at 366, which I will be happy with, 100mhz on FSB and 900 core.

I did raise my nForce MCP from 1.50v to 1.525v.
Can someone explain what nForce SPP/MPC does?
Also what is HT nForce SPP <->MCP?

Thanks again,
Michael
 
The SPP is the southbridge. The SPP <-> MCP thing is the interconnect between the northbridge and southbridge. You shouldn't need to change any settings related to the SPP since the CPU and RAM are handled by the northbridge.
 
You must log in or register to reply here.
Back
Top