Saga #2 - Q9450 @ 3.2Ghz woes

[Nova Phoenix]

Re: Saga #2 - Q9450 @ 3.2Ghz woes

Well if you think LLC was needed then all you really need to do is disable it and add more vcore.

Skews, ya you can try any setting for stability for sure and it may help or may not

Yes Vtt is FSB Voltage

You can try those gtlref settings, try +3 on 1 and -3 on the other

If the current settings with LLC enabled BSOD or RSOD at some point, then I'll disable it, increase the Vcore, and lower FSB OV. Although, at times before, doing both of those things hasn't helped for long, as the system eventually BSODs whilst idle mostly, or rarely whilst playing a 3D game.

As for GMCH OV, I think it is at either +0.200v or +0.300v. I don't know if the latter is really dangerous. My NB cooler is the motherboard's one - a big copper block thingy.

Currently, I have the system right where I want it. It's been on for almost 2 days straight with no problems with the current settings. Not saying they are the be-all-end-all to my woes of days past - not yet anyway.

Also, they have survived o/n tests from Orthos, Prime95 and a quick one from IntelBurnTest today (although that made the cores breach 70oC for a short time, I stopped the test when I saw those). Then again, tests were never really the problem (system always passed them, no errors anywhere).

EDIT: Got a BSOD whilst watching a Mugen demo (from the app, not youtube). Restared with the Vcore/LLC/Skew suggestions you gave me, lsdmeasap. I brought the NB down to +0.250v as well, and FSB down to +0.10v.

[Nova Phoenix]

Re: Saga #2 - Q9450 @ 3.2Ghz woes

Got an RSOD whilst playing Crysis Wars.

I'm beginning to think my CPU is f-ed. No matter what BIOS settings I try, it always ends up going kaput! Dunno what else to do people.

I'm waiting until I have enough money to get a system similar to the one in Gigabyte's "Core i7 920 at 4Ghz on air" PDF article. That won't be until next year when Nvidia's GTX285 is released though.

[Nova Phoenix]

Re: Saga #2 - Q9450 @ 3.2Ghz woes

It went BSOD again during a Nero Recode of a DVD!!

As an experiment, I decided to go back to my motherboard's original BIOS - version F3 (the one it had when I bought it in June). This one explicitly states that it features OC improvements. I thought it couldn't hurt to try it, so I Q-Flashed it. Now, here are my settings:

Robust Graphics Booster___________ [Auto]
CPU Clock Ratio ____________ [8]
Fine Clock Ratio ____________[0.5] << despite this, it's still 3.2Ghz
CPU Host Clock Control_ [Enabled]
CPU Host Frequency (MHz)__________ [400]
PCI Express Frequency (Mhz)_______ [100]
C.I.A. 2__________________________ [Disabled]
System Memory Multiplier (SPD)____ [2.00D]
DRAM Timing Selectable_______ SPD __ [Manual]

Performance Enhance = [Standard]

CAS Latency Time________________ 4
Dram RAS# to CAS# Delay_________ 4
Dram RAS# Precharge Delay_______4
Precharge Delay (tRAS)__________ 12

ACT to ACT Delay (tRRD)_________ 3
Rank Write to READ Delay________ 3
Write to Precharge Delay_________6
Refresh to ACT Delay______________42
Read to Precharge Delay__________ 3
Static tRead Value_______________8
Static tRead Phase Adjust________ [Auto]

CPU/PCIEX Clock Driving Control [800mV]
CPU Clock Skew Control [normal]
GMCH Clock Skew Control [Normal]

System Voltage Control____ [Manual]
DDR2 OverVoltage Control__ [+0.300V]
PCI-E OverVoltage Control_ [+0.1V]
FSB OverVoltage Control___ [+0.1V]
(G)MCH OverVoltage Control [+0.125V]
Loadline Calibration - [Disabled]
CPU Voltage Control_______ [1.25]

Limit CPUID Max. to 3.....................: [Disabled]
No-Execute Memory Protect............: [Enabled]
CPU Enhanced Halt (C1E)................: [Disabled]
CPU Thermal Monitor 2(TM2) ...........: [Disabled]
CPU EIST Function.........................: [Disabled]
Virtualization Technology................: [Disabled]

Update: Failed! Probably the Vcore's fault! I'm gonna boost it to the max (within the safety range of the chip of course) and see what happens then.

[Nova Phoenix]

Re: Saga #2 - Q9450 @ 3.2Ghz woes

It went BSOD again during another encode attempt.

Vcore is still up high (1.3680v in BIOS) and NB/GMCH is boosted to +0.300v. Other voltages are still at +0.10v (mem is still at +0.30v).

EDIT: After reading an Anandtech article on quad-OCing, I decided to test my VTT/FSB at +0.20v (article said this is okay for an fsb of 400mhz). I also read your responses, lsdmeasap, to someone who had a problem with crossfire, and I amped my NB to +0.300v and my PCIE voltage to +0.20v.

To test further, I lowered my Vcore below the Vid of 1.25, down to about 1.23 (in bios). In Windows, and during an encode I'm doing at this time, the Vcore is fluctuating between 1.200v and 1.216v. I'm surprised that it even POSTed let alone got into Windows.

This could be a good thing though. If it BSODs during the current encode, could that mean that it isn't the Vcore's fault? Previously I haven't set the BIOS Vcore below the VID - it's always been at it or above it (not counting vdrop/vdroop).

[Nova Phoenix]

Re: Saga #2 - Q9450 @ 3.2Ghz woes

Update: Below VID Vcore failed. Gonna boost Vcore to the safe-max again, leaving hte other voltages unchanged, and see what happens.

EDIT#1: Sorry for the continuous posting. To test for possible stability issues, I removed one of my HD3870X2s and switched out the memory for chips that haven't been used in ages (they were back when I had 8GB). BIOS settings remain pretty much the same as my last attempt, except Vcore (1.28750v in BIOS) and Loadline Calibration is Auto.

[Nova Phoenix]

Re: Saga #2 - Q9450 @ 3.2Ghz woes

The CPU is failing to be stable even at stock everything. Multiple tries of MeGui and RipBot have all failed, no matter how under-/over-clocked the CPU is or isn't. So my conclusion is is that my CPU is "dieing".

I have two options: tough it out (lol) until I have money for my x58 system, or buy a dual-core (my eyes are on the e8600) to tide me over until I have enough for the x58 rig.

Thanks for your help everyone, especially lsdmeasap. But it seems that all of your efforts were in vain. Sorry. :(