i7 920 BSOD OC questions

[Bloodstriker]

I ran a Real Temp sensor test last night and I got the dreaded BSOD. Real Temp is setup to check temperatures at load with Prime 95 (I used the blended test).

I find this really strange, I have been running the test at current settings everyday for almost a week now. I've even been okay with Prime 95 running for 7.5 hours straight.

I don't think I have any problems with core temperatures - Idle 36-39, Load 65.

My BIOS settings are as follows:

A.I. - X.M.P. Profile 2 (1.5v)
CPU Multiplier - Auto
BCLK - 169
DRAM @ 1352
CPU Voltage - 1.225
CPU PLL Voltage - 1.80
QPI/DRAM Voltage - 1.225
DRAM Bus Voltgage - 1.50

My system:

i7 920 C0/C1
Asus P6T Deluxe v1
3 x 2GB Patriot Viper Extreme 1333

I'm wondering if this has to with QPI/DRAM Voltage - should I be bumping that up a bit and leaving CPU Voltage the same?

I notice theres a bit of stuttering on the system - i.e. sometimes when I go to close a window or click on something, it seems to take a bit longer than usual for the system to respond.

 

[imported_Shaq]

Use LinX to test your CPU. It will show instabilities quicker. Run that and up CPU or QPI if it fails.

 

[Bloodstriker]

Originally posted by: Shaq
Use LinX to test your CPU. It will show instabilities quicker. Run that and up CPU or QPI if it fails.

Thanks for the advice Shaq.

I ran LinX with Memory to use set to 'All' and ran 85 times (4h 56m) without any problems. Max temp was 69C.

I'm guessing the BSOD was a one-off.

 

[Tullphan]

Originally posted by: Bloodstriker

Originally posted by: Shaq
Use LinX to test your CPU. It will show instabilities quicker. Run that and up CPU or QPI if it fails.

Thanks for the advice Shaq.

I ran LinX with Memory to use set to 'All' and ran 85 times (4h 56m) without any problems. Max temp was 69C.

I'm guessing the BSOD was a one-off.

Running it using "All" for memory, + 85 times w/no problems is a good thing!
I just ran mine 25 times using 1/2 the memory & it did ok. Earlier I had tried it using the max memory at a higher overclock than what I have now & got the BSOD...& I had it checked to stop on errors too.
Anyhow, you should be good to go!

 

[evolucion8]

If you run Linpack with lots of memory will be the best way to test QPI stability, but the CPU will work slighly less compared to memory minimum usage like 25% or less which is nice to test CPU overclock stability.

 

[imported_Arkhon]

Hey guys

I was wondering if you guys could help me i have a i7 920 and i would like to overclock it to 3.0 but i cannt seem to get it to run stable at a base clock 143x21=3.002.9 clock rate. I use prime95 for stability test at 2.9 it ran for three hours with no errors I am using turboV to overclock was wondering if you guys could help thanks

 

[imported_Shaq]

At 3.0 you shouldn't have to adjust anything for it to be stable. It might be the memory. Lower the memory multiplier to 6 and keep the timings on auto and see if it passes.

 

[imported_Arkhon]

thank you for the reply Shaq

Ok I was thinking the samething but i am not sure how to downclock my memory mulitiplyer because the i7 920 mulitiplyer is locked if you could tell me how or point me in the right deriction that would be great thanks for your help

 

[kyotousa]

Is it worth it to get the DDR3 1600 8-8-8-24 stick than the DDR3 1333 7-7-7 stick?

In term of real world experience & overclocking ability

 

[AdamK47]

I used settings I thought were stable before too by simply running Prime95 alone. I found by running LinX, MemTest, and Prime95 at the same time it uncovered problems (BSODs, errors) that running one app alone could not find. Screw the idiots who say they steal clock cycles from each other. This test works!