System Settings?

[Raife]

Long - But Please Help!

 

First, my BIOS settings

 

BIOS Version: P06

**System **

CPU Freq, MHz: 2666.7 GHZ

FSB Reference Clock, MHz: 1333.3

CPU Multiplier: [8X]

PCle x16-1 & 16-2, MHz: [Auto] - 100

PCle x16-3, MHz: [Auto] - 100

SPPc->MCP Ref Clock, MHz: [Auto] - 200

nForce SPP --> nForce MCP: [5X]

nForce SPP c-- nForce MCP: [5X]

**Spread Spectrum**

CPU Spread Spectrum: [Center Spread]

HT Spread Spectrum: [Auto]

PCle Spread Spectrurn(SPP): [Disabled]

PCle Spread Spectrurn(MCP): [Auto]

SATA Spread Spectrum: [Disabled]

**FSB & Memorv Config**

FSB - Memory Clock Mode: [unlinked]

FSB - Memory Ratio:

FSB (QDR), MHz: [1333]

Actual FSB (QDR), MHz: 1333.3

x MEM (DDR), MHz: 1600

Actual MEM (DDR), MHz: 1600

P1: [AUTO]

P2: [AUTO]

**Memory Timing Setting**

Memory Timing Setting: [Expert]

tCL (CAS Latency): [7]

tRCD: [7]

tRP: [7]

tRAS: [18]

Command Per Clock: [2T]

** Advanced Memory Settings **

tRRD: [Auto (5)] - 4

tRC: [Auto(40)] - 33

tWR: [Auto(12)] - 12

tWTR: [Auto(17)] - 16

tFAW: [Auto(25)] - 21

tREF: [Auto] - 7.7uS

**System Voltages**

CPU Core: [Auto] - 1.14V

CPU FSB: [Auto] - 1.1V

Memory: [1.900V] - 1.900V

nForce SPP: [Auto] - 1.30V

nForce MCP: [Auto] - 1.500V

GTLVREF Lane 0: [Auto] - -30mv

GTLVREF Lane 1: [Auto] - +10mv

GTLVREF Lane 2: [Auto] - +00mv

GTLVREF Lane 3: [Auto] - +00mv

**CPU Configuration**

Limit CPUlD Maxval: Disabled

Intel Speedstep: Disabled

xPPM Mode: [Native]

CPU Thermal Control: [Disabled]

C1E Enhanced Halt State: [Enabled]

Execute Disable Bit: [Enabled]

Virtualization Technology: [Enabled]

CPU Core 0: [Enabled]

CPU Core 1: [Enabled]

CPU Core 2: [Enabled]

CPU Core 3: [Enabled]

 

Red text above is the only current changes I have in the bios, everything else is default. The changes are based on my memory G.Skill DDR3 PC3-12800 (2x2GB) Model F3-12800CL7D. The default values in the bios are 9-9-9-24-2T, the manufacture setting is 7-7-7-18-2T and suggested voltage of 1.9V

 

So here is the situation. I built the computer in my sig about 3 months ago and have not had more than a couple hours of stability since then. Generally what happens is the computer boots and shortly afterwards I begin to get video artifacts (jaggies or

[Crazy_Nate]

What memory speed is the standard for the motherboard?

 

Try switching from unlinked to linked, with a divider to get you to run at that speed. Give it the maximum recommended RAM voltage (1.9V?).

 

If that doesn't work and you still get memtest errors...loosen up your advanced timings and try again...then loosen up your primary timings. Once you get some settings that work, you can go from there.

 

I would test in this order: memtest, prime, 3DMark. Don't worry about drivers until you've gotten the first two figured out.

 

Good luck! :thumbs-up:

 

Edit: If you're still having video trouble, try using just one graphics card initially...get it stable, then add the second.

[Raife]

The manual says "Support dual channel JEDEC DDR3-1333 and SLI-Ready memory up to 2000MHz. Supports up to 8 GBs DDR3 memories"

 

On the G.Skill website it states for my memory set "*G.Skill guarantees 2 Dimms Dual channel operation would reach announced specification."

 

I have run the memory linked and at the same speed as the FSB with and without the tighter timing - both ways I still get the artifacts.

 

I'll start playing with some of the other settings.

[Crazy_Nate]

The manual says "Support dual channel JEDEC DDR3-1333 and SLI-Ready memory up to 2000MHz. Supports up to 8 GBs DDR3 memories"

 

On the G.Skill website it states for my memory set "*G.Skill guarantees 2 Dimms Dual channel operation would reach announced specification."

 

I have run the memory linked and at the same speed as the FSB with and without the tighter timing - both ways I still get the artifacts.

 

I'll start playing with some of the other settings.

 

Has it passed memtest86 at those settings? I would work on the memory first, then check the CPU, then the graphics cards (which might end up being the problem - test them one at a time).

 

I would expect video problems in BIOS to be related to video cards.

[Raife]

Well last night the rig ran for several hours w/o crashing. I was running 1 stick of RAM and one video card. Also changed the following BIOS settings.

 

CPU Spread Spectrum: [Center Spread] switched to disabled

HT Spread Spectrum: [Auto] switched to disabled

PCle Spread Spectrurn(MCP): [Auto] switched to disabled

FSB - Memory Clock Mode: [unlinked] switched to linked

x MEM (DDR), MHz: 1600 auto set it at 1333

Memory Timing Setting: [Expert] switched to auto

tCL (CAS Latency): [7] auto set it at 9

tRCD: [7] auto set it at 9

tRP: [7] auto set it at 9

tRAS: [18] auto set it at 24

Command Per Clock: [2T] auto set it at 1T

Memory: [1.900V] - 1.900V (kept at 1.900V rather than going back to default)

C1E Enhanced Halt State: [Enabled] to disabled

Execute Disable Bit: [Enabled] to disabled

Virtualization Technology: [Enabled] to disabled

 

Then ran memtest on the first stick overnight. On 23 passes received errors on pass 14 and 17. The second stick is running at home now.

[Raife]

So I rechecked the EVGA website today for compatable RAM and a set of sticks that are nearly identical to mine (I think the heatspreader is the only difference) were listed as compatable

 

G.SKILL F3-12800CL7D-4GBP1 <- The sticks listed as compatable

G.SKILL F3-12800CL7D-4GBHZ <- My sticks

 

Does this mean mine are or are not compatable?

[Raife]

So the rig ran stable at idle overnight on 1 video card and 2 mem sticks. I ran both stick is in slot 0 and 2 and received a couple of errors each

 

 

 

Stick 1 - Slot 0, error on pass 14 and 19 of 23 passes

 

Stick 1 - Slot 2, error on pass 2 of 20 passes

 

Stick 2 - Slot 0, error on pass 1 of 28 passes

 

Stick 2 - Slot 2, error on pass 6 of 42 passes

 

 

 

So memory errors have gone down (I know I not supposed to get any, but its improving which leads me to believe further BIOS tweaking should get it)

 

Now testing 2 stick at the same time in slots 0 and 2

Edited August 13, 2008 by Raife

[Raife]

Ok - so I have continued to test components individually and here is where things stand.

 

 

 

RAM

 

On stock SSP voltage my single stick RAM was getting 1-2 errors in 20-42 passes

 

 

 

Then tried 2 sticks in dual channel and got 7 errors out of 19 passes on stock SSP voltage and 7 errors out of 32 passes on with a SSP voltage of 1.4V

 

 

 

Upped the SSP voltage to 1.45V and got 4 errors on 39 passes.

 

 

 

So I am wondering what I can safely run my SSP voltage at to see if more juice is the issue.

 

 

 

Video Cards

 

So each card seems to run fine by itself in the slot closest to the CPU. I thought maybe once I switched the cards I would begin to have artifacting agian, but not so far.

 

 

 

So what are my next steps? The RAM is still running default timings and speed, but I have upped the voltage to the manufacture recommended setting (1.9V). The cards are running singlarly rather than SLI.

 

 

 

Thoughts?

[Raife]

Ok, I stand corrected - last night while running memtest with the SPP at 1.5 I recieved video artifacts during memtest. I also had a hardlock while leaving the system running overnight (fully booted). Both of these instances were with the second video card by itself. Therefore I think the second video card is goin to be RMA'd.