GA-EP45-UD3P BIOS settings verify

[ozczecho]

Hi guys,

I am a long time lurker but finally decided to join up. Anyways I have a:

GA-EP45-UD3P
2 * 2G DDR2 RAM Corsair twin2x4096-8500c5d
Intel E8500 cpu

Below are my BIOS settings. My PC is only slightly tweaked. I didnt want to go overboard. The system runs stable for about 2 months (this includes 24hr memtest runs) and then suddenly after 2 months BSoD and memtest fails. I have returned the corsair ram twice. I now have a 3rd set and am wondering if my BIOS settings are OK for my setup.

Any help would be greatly appreciated.

Robust Graphics Booster.........................[ Auto     ]
CPU Clock Ratio.................................[ 8  x      ]
Fine CPU Clock Ratio............................[ + .0    ]
CPU Frequency 4GHz............................[ 500 x 8  ]
 
******Clock Chip Control******
 
CPU Host Clock Control..........................[ Enabled  ]
CPU Host Frequency (Mhz)........................[ 500     ]
PCI Express Frequency (Mhz).....................[ 100      ]
C.I.A 2 [Disabled] 
 
 
******Advanced Clock Control******

CPU Clock Drive..........................[800mV]
PCI Express Clock Drive................[900mV]
CPU Clock Skew..........................[0ps]
MCH Clock Skew..........................[0ps]

******DRAM Performance Control******
 
Performance Enhance.............................[ Standard    ]
Extreme memory profile.......................[ disabled ]
MCH Frequency Latch.........................[ 333   ]
System Memory Multipler.........................[ 2.00B    ]
Memory Frequency 1066............................[ 1000      ]
DRAM Timing Selectable..........................[ Manual   ]
 
CAS Latency Time................................[ 5        ]
tRCD............................................[ 5        ]
tRP.............................................[ 5        ]
tRAS............................................[ 15      ]
 
******Motherboard Voltage Control******
 
Load-Line Calibration...........................[ Disabled ]
CPU Vcore         [ 1.25 ]...................[ 1.28125v    ]
CPU Termination   [ 1.20    ]...................[ 1.20v   ]
CPU PLL           [ 1.50    ]...................[ 1.50v    ]
CPU Reference     [ 0.760   ]...................[ .815V  ]
 
MCH Core          [ 1.100   ]...................[ 1.220v    ]
MCH Reference     [ 0.760   ]...................[ .70v   ]
MCH/DRAM Reference [ 0.900   ]...................[ auto  ]
ICH I/O           [ 1.500   ]...................[ auto   ]
ICH Core          [ 1.100   ]...................[ auto   ]

DRAM Voltage     [ 1.800v]...................[2.100v]
DRAM Termination [0.900v]...................[auto]
Channel A Reference [0.900v]................[auto]
Channel B Reference [0.900v]................[auto]

[Chike]

I guess some concider 500x8 a slight tweak LOL.
Have you tried memtest with one stick only in both slots seperately with default settings other than those needed for DDR-1066?

[ozczecho]

Thanks Chike for the reply. The thing is the systems runs reliably for about 2 months, used every day, on or in standby mode 24*7. Then without warning BSoD and then when I run memtest CD, red fills the screen...

This has happened twice so far. In each case I had the RAM replaced (RMA). I now have a new replacement RAM (RAM number 3) and want to make sure that its not my settings that is killing the RAM.

Cheers

[Chike]

Yes I got that it's the 3rd pair, that's why I suggested testing the board itself. It should be able to run at least one of the two modules without errors in all slots.

[monza1412]

Do you have the same problems in this strap (400/2D)?
Did you try to raise your MCH a little?

[ozczecho]

OK Chike, I should have supplied more info.

After the first pair of RAM failed, I did a bit of testing.

* Reset BIOS and used the Optimised Defaults. System failed, Memtest red, primetest failed with rounding errors
* Removing 1 stick of RAM the system worked.
* I switched the RAM sticks around, and system still worked.
* I put the RAM in together again and system failed. Memtest red.
* I put the RAm into different slots (YELLOW) instead of (RED) no difference.

When I got the new RAM pair (the 2nd pair) I used the BIOS settings above and stress tested system using memtest and also primetest..both for 24hrs all up....no errors in either, temps OK. When this pair failed, again I tested with memtest after failure and red filled the screen. This time though it would not allow me to boot into Xp. The system was cactus. pretty much straight away. I didnt do any testing after that. Returned RAM under RMA and now got the 3rd pair.

I am worried that the BIOS setting above are killing my RAM - eventhough the settings look OK to me. But I am by no means an experts, hence I am here to see if I have screwed something up.

If the MoBo is faulty wouldn't the system fail straight away? As I mentioned, the system is rock solid for about 2 months. Then a BSoD and thats it, the end. Strange.

Again Thanks for ur help

[ozczecho]

Some screenshot with the current memory installed and hwMonitor. BIOS settings as above.

Cheers

Mike

[Chike]

Originally Posted by ozczecho

I am worried that the BIOS setting above are killing my RAM - eventhough the settings look OK to me. But I am by no means an experts, hence I am here to see if I have screwed something up.

Neither am I.

I thought the other two pairs failed right away I didn't realize they were working too for a while.

Can't help with the settings because I'm not familiar with all the references, maybe someone that is can.

[Psycho101]

With a DRAM voltage of 2.1V the references will all be auto set to 1.050V. This shouldn't be enough to kill the RAM. Timings themselves are also not really a thing that will threaten a components life. The two component killers are voltage and Heat.

have a look in the PC Health section of the BIOS and see what the BIOS reports for DRAM/DIMM voltage. This isn't exact but it's correct enough for diagnostic purposes. You may find that you are getting more voltage than needed. You could safely go down one nitch on VDIMM and likely be just as stable. Eg when I set 2.1V I get 2.092V in PC Health. If I set 2.12V then I get 2.12V. I just leave it at 2.092. Note that I have the same brand of sticks as you, running a slight overclock on them. You may not get the same results as me as the actual module design and components used varies even on the exact same model.

If you want to be more sure on RAM voltage and can use a DMM (Digital Multi Meter) then you could find the measuring points for your board via a Google search, and see exactly what they are receiving.

Run a stress test on the system, something like Prime 95 on Large FFT. After 30 minutes with a closed case, remove the side panel, earth yourself to something metal, then tough the RAM. It will be warm, but you should be able to hold your hand on it almost indefinitely without discomfort. If it is getting excessively warm, ie you have to remove your hand in under 10 sec due to the heat, re check to be certain that voltage and RAM frequency are set OK.

If voltages and frequency are set fine, it may be a cooling issue. In my experience, it takes quite a bit of heat over a sustained period to kill RAM, but as every component varies, who knows the ideal for your sticks. In this instance, consider adding additional cooling. There are special RAM coolers available that clip on. they can be noisy, but some are RPM controllable. Others can be tuned down with a simple adapter. Other than that, a more powerful intake and extraction fan, and maybe a 120mm fan mounted blowing on to the RAM would be great. Personally I'd use the 120mm option over a dedicated RAM coolar as it'll move more air at lower volume level (in dB). Again, you can reduce the voltage and thus RPM of the fan to suit cooling and noise requirements.

I would also encourage you to run Memtest over night for at least 8 hours to verify that the new sticks are currently 100%. It's not often I recommend the use of Memtest, but in cases of hardware failure it's a priceless tool.

[ozczecho]

OK Psycho, thanks for that. I will try your suggestions and see.

I had a look @ the BIOS PC Health and as u said it reported 2.096. I will run the stress test on the weekend to see what temps I get....

Again thanks for your insight.